Greetings DMs (and all you players bad at obeying instructions),
I have in my party a Half-Elf Fighter 1/ Hexblade 10/Pact of the Blade, a Hill Dwarf Peace Cleric 11, Mark of Shadow Elf Rogue 8 Arcane Trickster/Warlock of the Deep 3, Shadar-Kai Fire Circle Druid 10/Grave Cleric 2, Human Fighter Battlemaster/Polearm Master Great Weapon Master. Oh yea, both Hexblade and Fighter have Sentinel.
This is a solid party who...MOST of the time work together quite well. Lots of damage curve and between the Peace Cleric and Fire Druid/Grave Cleric can either keep people up or get them up/stabilize them as a reaction, so they are difficult to down and keep down.
It is a very fun challenge to find opponents who are a threat but do not overwhelm the party and the presence of Eldritch Smite makes having a super strong BBEG less effective since the Hexblade can just prone them and once on the ground they get the blender treatment.
I am an experienced DM and have ways to handle things (environment, intelligent baddies, etc), but am just asking the DM Hivemind for suggestions to help give these guys a run for their money.
Most of the time the solution to challenging encounters at higher tiers involves just using overwhelming force that vastly exceeds the encounter budget, but I would note that I really don't see very many good ranged options in that group, so a swarm of ranged foes will be a pain -- 6x flameskull isn't even a deadly encounter -- as will creatures that punish melee, such as a remorhaz or two (as a bonus: swallow denies line of sight, which blocks most healing).
1) Add objectives to combat rather than just "kill all the monsters or the monsters kill you". For example keep somebody alive, kill the enemies before they raise the alarm or stop the bad guy from performing the ritual.
2) Add more monsters
3) Adjust the monster by giving them more HP or add some powers, there are some decent 3rd party books like Monster talents on the dmsguild that help.
There's also the well-known trick: "This creature is immune to the prone condition/can't be knocked over." You can massage that into a bunch of different thematic approaches, but always remember that with the right justification anything is possible on the DM's side.
Rollback Post to RevisionRollBack
I know what you're thinking: "In that flurry of blows, did he use all his ki points, or save one?" Well, are ya feeling lucky, punk?
Avoid making things arbitrarily overcome the character's abilities, it will feel like there was no point in them gaining their abilities.
Anything with Flyby would be effective against them. Also consider dropping monsters in at the most inopportune moments - for example, the party just lost control of a wagon and it's hanging off a cliff. Suddenly, the people on the road are attacked by decent level bandits, and the person hanging from the wagon notices a cave which contains a monster who's interested in this morsel of food.
Making the combat a puzzle (which you need to fight through or die) makes for a good combat. Introducing monsters which make them think is also good - I had a paladin who relied entirely on smite. I introduced (by chance, but it worked like I'd planned it) a monster which was immune to Radiant damage. The Paladin worked it out, and realised they would have to try new things to win this. That's the sort of thing you need to do.
+1 to the Remorhaz I was playing in a party awhile back where we were steam rolling everything. We were 5 level 9’s, we had a lot of fire damage in the party and 1 Remorhaz stopped us in our tracks and killed, dead, dead one of the PCs. One or two of these might do the trick for this party. But I’m guessing that might not fit thematically. I am not a fan of the idea of just throwing something at them because it can give them a run for their money.
Normally as a DM I would not even consider intentionally finding something to counter a players ability, if it was not organic to the setting, typically when I do counter a PC's ablity in my games it is completely by accident. I am Not a fan of DM meta gaming. But dealing with something like Eldritch Smite, which is a little op. Being able to prone a huge creature without any kind of a check/save on the creatures part at all is in my mind absurd and ridiculous, especially when the PC can do it every turn, and can only use a level one slot. If that PC is first in the initiative and the monster is last, with a party that can crank out serious damage in the middle…. I can see where this would be hard to present the party with a challenge.
For a BBEG, or a this just got real encounter, I would just use whatever you have planned, let them prone and pummel the bad guy, and then pull the old "this is not my true form", and have the bad guy transform/drop its disguise, boom it’s an Ancient Dragon. Ancient Deep Dragon fits the requirements nicely, it is gargantuan (so no proning with Eldritch Smite) has the ability to change its shape(RAW) and is only a CR 18, 201 hp. The legendary action Commanding Spores will negate the clerics using their reactions to heal/stabilize. It can use its tail attack to knock the PCs prone and give them some of their own medicine. Add in some lair actions, and some minions, maybe some kobolds/or some kind of dragonkin, with short bows or some other ranged attack. To top it off I would have the minions doing some kind of ritual that has to be stopped, even if it is just the old sacrifice the princess to the dragon. It would add another layer to the encounter on top of the oh $#!t it’s a dragon!
Just keep in mind that you may need to nerf your encounter mid combat if it turns out to be an overbuilt TPK by design.
Welcome to the DM nightmare...no not really a nightmare. It is the expected result of running at level 11 or 12 and allowing players to run ridiculous OP subclasses and builds.
You are going to have to ignore the CR values of monsters, and give your monsters many many Legendary Resistances and Actions. You will have to give them minions. And you will have to make sure the BBEG is airborne and launching death from over 100 feet above the group. Take an ancient Red Dragon, then tack on all the abilities and spells of say a 15th level Divine Soul Sorcerer, and that would be a start.And EVEN THEN, thanks to Sentinel, if a PC gets close, the fight is over.
Welcome to the DM nightmare...no not really a nightmare. It is the expected result of running at level 11 or 12 and allowing players to run ridiculous OP subclasses and builds.
You are going to have to ignore the CR values of monsters, and give your monsters many many Legendary Resistances and Actions. You will have to give them minions. And you will have to make sure the BBEG is airborne and launching death from over 100 feet above the group. Take an ancient Red Dragon, then tack on all the abilities and spells of say a 15th level Divine Soul Sorcerer, and that would be a start.And EVEN THEN, thanks to Sentinel, if a PC gets close, the fight is over.
Although I do agree that Eldritch Smite is OP, and even ridiculous, it really just needs a few minor tweaks to be brought back into the realm of reasonable. It should be reduced to large or smaller, or even medium or smaller and require the target to make a save, on a fail they get knocked prone.
Anyone who has DMed, 5e knowns that the CR system is a baseline and is not a hard line, also there is a lot of miss use, misunderstanding of how to balance an adventuring day, with encounters, short rest, long rest, ect, using the CR system. I usually find that people who are having issues with the CR system aren't using it correctly. and Yes, we have to remember it is based on 5e (kinder, gentler, DnD) not OD&D or 2e.
"Give your monsters many many Legendary Resistances and Actions. You will have to give them minions. And you will have to make sure the BBEG is airborne and launching death from over 100 feet above the group. Take an ancient Red Dragon, then tack on all the abilities and spells of say a 15th level Divine Soul Sorcerer, and that would be a start." that would not be a start that would be a TPK, and intentional one. That Sorcerer build alone if played right could TPK a level 11 party in like maybe one round, two at the most. That's without giving it legendary actions, legendary resistances, a breath weapon that does 26d6 fire damage, and 546 hit points.
Anyone who has DMed, 5e knowns that the CR system is a baseline and is not a hard line, also there is a lot of miss use, misunderstanding of how to balance an adventuring day, with encounters, short rest, long rest, ect, using the CR system. I usually find that people who are having issues with the CR system aren't using it correctly.
While it's true that a significant part of the problem with the CR system is that people don't really get the six encounter day, the fact that the CR system is based around a play style that no-one actually wants to use is itself a problem. However, independent of that, the CR system has a math problem: PC power scales up faster than monster power. A typical CR 20 has about 3.5x the damage and defense of a CR 5; a typical level 20 has something like 3x the offense and 8x the defense of a level 5, not to mention a likely boatload of weird powers that can end encounters in a way that doesn't directly interact with monster power. As a result, while a reasonable encounter at level 5 is something like 50% of the daily budget, a reasonable encounter at level 20 is more like 100%.
Firstly, DnD 5e is not design for players to die both in that it isn't very likely and that it isn't made fun. It doesn't have the rogue like elements of earlier editions where you have random characters and an xp scale that encourages catch up. I wouldn't recommend trying to kill players for that reason which means your tension needs to come from else where. Set goals that they can fail in ways other than the players dieing. For example give them rescue or capture alive missions. Even rats can be challenging if you are trying to stop them killing commoners as opposed to demigod players. You can go full plague tale on them.
Secondly If they are performing as a higher level it only makes sense that they will be asked to and may try more challenging tasks. Consider treating them as a party a couple of levels higher if you are going to make the challenge putting the players at risk.
Thirdly there are ways to put players at risk other than just monsters and you should try a variety of different things to keep it interest and avoid turning it into a tit v.s tat of optimizing. Use things like:
Environmental hazards can add challenge to the party and change play style. They do not lose threat with hit points like a group monsters does and it's easier to justify non combat solutions. This can be dungeons filled with poison miasmas that drain health constantly, flooded underground lairs, tight corridors that require squeezing, trap laden temples, strange magical pocket dimensions and unusual natural features.
Finally there are elements of encounter design that can reduce the effects of certain strategies.
Breaking the health bar up into pieces reduces player damage because damage that exceeds a block is lost. This can be done wither in a single monster by giving it damage gates like boss stages or by using multiple monsters. Though keep in mind that having more targets also increases AOE damage.
Special defeat conditions can shift the goal of the battle. This can be disabling devices that strengthen the boss or weaken the party, killing a healer first, needing to hit some specific weakness of the boss or needing to attack the boss at specific times. If you want a specific strategy or player to shine you can even make it so that a monster can only die a specific way e.g the final blow must be from radiant damage.
Damage modifiers allow you to change how certain attack effect certain monsters. They should be varied so the same kind of attacks aren't always used and players don't fall into too much of a pattern. Legendary resistances are an example of damage modifiers in the form of a monster resisting until x failed saves. You can conceivably have the same thing for melee attacks or any other kind of attack. As well as what they effect you can also change when they kick in for example a monster may regenerate half of damage it has taken at the end of a turn or on it's turn. There are also all of the usual resistances and immunities as well as special protections like enemies having disadvantage on attacks in darkness for example.
Disabling effects can reduce the relevance of hit point healing. I personally dislike disabling effects that take players out of combat immediately but it's reasonable for penalties to come into play when players hit 0. Powerful monsters can have rider effects that come into player when players hit 0 like possession, summoning more enemies, things like flesh to stone which give a player a timer to death like normal bleeding out would, or effects like stat drain which weaken a player temporarily.
Anyone who has DMed, 5e knowns that the CR system is a baseline and is not a hard line, also there is a lot of miss use, misunderstanding of how to balance an adventuring day, with encounters, short rest, long rest, ect, using the CR system. I usually find that people who are having issues with the CR system aren't using it correctly.
While it's true that a significant part of the problem with the CR system is that people don't really get the six encounter day, the fact that the CR system is based around a play style that no-one actually wants to use is itself a problem. However, independent of that, the CR system has a math problem: PC power scales up faster than monster power. A typical CR 20 has about 3.5x the damage and defense of a CR 5; a typical level 20 has something like 3x the offense and 8x the defense of a level 5, not to mention a likely boatload of weird powers that can end encounters in a way that doesn't directly interact with monster power. As a result, while a reasonable encounter at level 5 is something like 50% of the daily budget, a reasonable encounter at level 20 is more like 100%.
For sure! I don't know any DMs that run 6 CR based/combat encounters in an adventuring day, and I don't know any players that would want it that way. But as a result we end up with people running one or two easy encounters, or they skip the easy and do one or two medium encounter, then one hard or deadly, and let the players take short or long rest in-between, and wonder why the party runs all over everything.
As far as the math problem. I think, although there may very well be some mathematical issues, I think it is just as much a perception issue. Yes players level/scale faster, but there is zero correlation between player level and CR as far as as level 5 PC has nothing to do with a CR 5 monster, and no correlation between CR and PC level should be made. Also saying that CR20 should be 15x more than a CR5 was is not how the CR system works and it can't work that way. If that is how the CR system worked all CR10 enemies would be 100% identical with pasted on names and themes. In a system which worked that way a CR15 Dragon and a CR 15 Archfiend would have the exact same health, damage output, and abilities, and that would be boring. Instead the CR system is a baseline system which takes averages of monsters and categorizes them for the ease of selecting what the average difficulty will be against an average party with X number of members. That's how you end up with two CR12 monsters with sometimes extremely varied hp, damage output, and abilities. A monster with low HP and high damage might be the same CR as a monster with high hp and lower Damage, depending on the other ability's of the monster.
As far as the math problem. I think, although there may very well be some mathematical issues, I think it is just as much a perception issue. Yes players level/scale faster, but there is zero correlation between player level and CR as far as as level 5 PC has nothing to do with a CR 5 monster, and no correlation between CR and PC level should be made.
The concept behind CR is that a CR x monster is a medium encounter for a level x party of 4 -- level 5 party gets 1xCR 5, level 20 party gets 1xCR 20 (the math is slightly erratic, but it's mostly true), and if you look at the standards for CR x monsters, it turns out that things like AC are tuned to match the expected capabilities of level x PCs.
Also saying that CR20 should be 15x more than a CR5 was is not how the CR system works and it can't work that way. If that is how the CR system worked all CR10 enemies would be 100% identical with pasted on names and themes.
The CR system is designed around expected damage output and hit points. There are some anomalous monsters, but for the most part you can assert "A CR x monster has (x+1)*15 hp and does (x+1) damage per round, with an xp value of (x+1)*(x+1)*50, possibly with some tradeoff of being above-CR on one measure and below on the other" and be ballpark correct up through CR 20.
As far as the math problem. I think, although there may very well be some mathematical issues, I think it is just as much a perception issue. Yes players level/scale faster, but there is zero correlation between player level and CR as far as as level 5 PC has nothing to do with a CR 5 monster, and no correlation between CR and PC level should be made.
The concept behind CR is that a CR x monster is a medium encounter for a level x party of 4 -- level 5 party gets 1xCR 5, level 20 party gets 1xCR 20 (the math is slightly erratic, but it's mostly true), and if you look at the standards for CR x monsters, it turns out that things like AC are tuned to match the expected capabilities of level x PCs.
Also saying that CR20 should be 15x more than a CR5 was is not how the CR system works and it can't work that way. If that is how the CR system worked all CR10 enemies would be 100% identical with pasted on names and themes.
The CR system is designed around expected damage output and hit points. There are some anomalous monsters, but for the most part you can assert "A CR x monster has (x+1)*15 hp and does (x+1) damage per round, with an xp value of (x+1)*(x+1)*50, possibly with some tradeoff of being above-CR on one measure and below on the other" and be ballpark correct up through CR 20.
That may very well be the intended design concept( I have no idea), but if you go into encounter builder right now and set up a 4 member level 5 party and then filter so you only see level 5 monsters, you get easy encounters, not medium, With a 4 member level 5 party CR6-CR7 is medium, CR8 is hard, and CR9-whatever is deadly, there should be a feature that says impossible so you know....but is doesn't, I'm guessing anything above a CR11 or CR12 would be impossible for a 4 member level 5 part. So if you filter the encounter builder for CR5 versus a level 5 party of for 4, and you look at the AC for the first 25 monsters the AC ranges from 9 to 19 with no meaningful average. In the same scenario you get hit points from 67 to 135, that's a 68 point swing. With HP you at least have an average of 90-114 ish 10 out of 25 were in this range. This whole concept completely breaks down at CR1 CR2 a party of 4 level 1 Anythings against some of the CR-1 monsters PC's are gonna die. Same goes for CR2 at these CR levels, when its already hard not to accidently TPK, I would not trust the CR system at all CR1-3.
The expected damage out put and hit points doesn't really gel either. With the HP 90ish is the low end of the average which adds up with your math (5+1)*15 =90 hp. But in the same 25 CR5 sample the average was over 15 points of damage, that was assuming they hit 50% of the time. (x+1) damage in this sample should be (5+1) =6 damage per round. The average was more then double that (in the small sample I did), but there was not a single monster in the sample that only did (x+1) damage per round.
Realistically none of this matters because if you are running a party of 4 level 5 Hill Dwarf barbarians. All of these CR 5 monsters would be toast. Conversely a party of 4 level 5 Wizards doing average damage of their highest level spell slots might take out anything in the 90-114 hit point category before any of them die, but all the wizards would need to go first, and the CR5 monster would need to fail all of its saves. In this scenario there would be a dead wizard by the CR5 monster's second turn, making this easy encounter a deadly one.
All of that was my longwinded way of saying the CR system is still just a base line. Its a good way to gage if your homebrew monsters are in the ball park for your party. But ultimately you have to know your party and what they are capable of. In my experiences the CR system usually errors in favor of going easy on the party at least CR level 5-15ish, before that it seems like it errors on the side of killing the PCs.
The expected damage out put and hit points doesn't really gel either. With the HP 90ish is the low end of the average which adds up with your math (5+1)*15 =90 hp. But in the same 25 CR5 sample the average was over 15 points of damage, that was assuming they hit 50% of the time. (x+1) damage in this sample should be (5+1) =6 damage per round. The average was more then double that (in the small sample I did), but there was not a single monster in the sample that only did (x+1) damage per round.
The expected damage out put and hit points doesn't really gel either. With the HP 90ish is the low end of the average which adds up with your math (5+1)*15 =90 hp. But in the same 25 CR5 sample the average was over 15 points of damage, that was assuming they hit 50% of the time. (x+1) damage in this sample should be (5+1) =6 damage per round. The average was more then double that (in the small sample I did), but there was not a single monster in the sample that only did (x+1) damage per round.
If you ignore a few of the bad assumptions Cr seems to make it does describe monsters fairly accurately, the issue is that there is a huge range in the power of player builds. This is partly due to what 5e considers optional rules because encounter balance doesn't take into account things like feats, multiclassing and magic items which many parties treat as standard. Its also out dated because the newer classes and subclasses tend to be stronger than many of the launch options.
If you really have no clue of where the party should be then you should be able to match it to the Create a monster table. Average the hit chance and ac and match it to the table, for health and damage compare the sum of the parties health and damage to the table. You can then treat the party as that level and CR should work a little better. However I think you inevitably reach a level where you need to be more reactive and just adjust the difficulty up and down based on how encounters go.
I also think that by the time you reach level 10 encounters need to start being more than just combat. Legendary monster like vampires become easily defeatable so unless you are going to throw demigods at them on the regular the goals need to be more than just simply killing them.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
Greetings DMs (and all you players bad at obeying instructions),
I have in my party a Half-Elf Fighter 1/ Hexblade 10/Pact of the Blade, a Hill Dwarf Peace Cleric 11, Mark of Shadow Elf Rogue 8 Arcane Trickster/Warlock of the Deep 3, Shadar-Kai Fire Circle Druid 10/Grave Cleric 2, Human Fighter Battlemaster/Polearm Master Great Weapon Master. Oh yea, both Hexblade and Fighter have Sentinel.
This is a solid party who...MOST of the time work together quite well. Lots of damage curve and between the Peace Cleric and Fire Druid/Grave Cleric can either keep people up or get them up/stabilize them as a reaction, so they are difficult to down and keep down.
It is a very fun challenge to find opponents who are a threat but do not overwhelm the party and the presence of Eldritch Smite makes having a super strong BBEG less effective since the Hexblade can just prone them and once on the ground they get the blender treatment.
I am an experienced DM and have ways to handle things (environment, intelligent baddies, etc), but am just asking the DM Hivemind for suggestions to help give these guys a run for their money.
Thanks,
Evan
Most of the time the solution to challenging encounters at higher tiers involves just using overwhelming force that vastly exceeds the encounter budget, but I would note that I really don't see very many good ranged options in that group, so a swarm of ranged foes will be a pain -- 6x flameskull isn't even a deadly encounter -- as will creatures that punish melee, such as a remorhaz or two (as a bonus: swallow denies line of sight, which blocks most healing).
You have all the tools in your toolkit already.
1) Add objectives to combat rather than just "kill all the monsters or the monsters kill you". For example keep somebody alive, kill the enemies before they raise the alarm or stop the bad guy from performing the ritual.
2) Add more monsters
3) Adjust the monster by giving them more HP or add some powers, there are some decent 3rd party books like Monster talents on the dmsguild that help.
There's also the well-known trick: "This creature is immune to the prone condition/can't be knocked over." You can massage that into a bunch of different thematic approaches, but always remember that with the right justification anything is possible on the DM's side.
I know what you're thinking: "In that flurry of blows, did he use all his ki points, or save one?" Well, are ya feeling lucky, punk?
Avoid making things arbitrarily overcome the character's abilities, it will feel like there was no point in them gaining their abilities.
Anything with Flyby would be effective against them. Also consider dropping monsters in at the most inopportune moments - for example, the party just lost control of a wagon and it's hanging off a cliff. Suddenly, the people on the road are attacked by decent level bandits, and the person hanging from the wagon notices a cave which contains a monster who's interested in this morsel of food.
Making the combat a puzzle (which you need to fight through or die) makes for a good combat. Introducing monsters which make them think is also good - I had a paladin who relied entirely on smite. I introduced (by chance, but it worked like I'd planned it) a monster which was immune to Radiant damage. The Paladin worked it out, and realised they would have to try new things to win this. That's the sort of thing you need to do.
Make your Artificer work with any other class with 174 Multiclassing Feats for your Artificer Multiclass Character!
DM's Guild Releases on This Thread Or check them all out on DMs Guild!
DrivethruRPG Releases on This Thread - latest release: My Character is a Werewolf: balanced rules for Lycanthropy!
I have started discussing/reviewing 3rd party D&D content on Substack - stay tuned for semi-regular posts!
+1 to the Remorhaz I was playing in a party awhile back where we were steam rolling everything. We were 5 level 9’s, we had a lot of fire damage in the party and 1 Remorhaz stopped us in our tracks and killed, dead, dead one of the PCs. One or two of these might do the trick for this party. But I’m guessing that might not fit thematically. I am not a fan of the idea of just throwing something at them because it can give them a run for their money.
Normally as a DM I would not even consider intentionally finding something to counter a players ability, if it was not organic to the setting, typically when I do counter a PC's ablity in my games it is completely by accident. I am Not a fan of DM meta gaming. But dealing with something like Eldritch Smite, which is a little op. Being able to prone a huge creature without any kind of a check/save on the creatures part at all is in my mind absurd and ridiculous, especially when the PC can do it every turn, and can only use a level one slot. If that PC is first in the initiative and the monster is last, with a party that can crank out serious damage in the middle…. I can see where this would be hard to present the party with a challenge.
For a BBEG, or a this just got real encounter, I would just use whatever you have planned, let them prone and pummel the bad guy, and then pull the old "this is not my true form", and have the bad guy transform/drop its disguise, boom it’s an Ancient Dragon. Ancient Deep Dragon fits the requirements nicely, it is gargantuan (so no proning with Eldritch Smite) has the ability to change its shape(RAW) and is only a CR 18, 201 hp. The legendary action Commanding Spores will negate the clerics using their reactions to heal/stabilize. It can use its tail attack to knock the PCs prone and give them some of their own medicine. Add in some lair actions, and some minions, maybe some kobolds/or some kind of dragonkin, with short bows or some other ranged attack. To top it off I would have the minions doing some kind of ritual that has to be stopped, even if it is just the old sacrifice the princess to the dragon. It would add another layer to the encounter on top of the oh $#!t it’s a dragon!
Just keep in mind that you may need to nerf your encounter mid combat if it turns out to be an overbuilt TPK by design.
Welcome to the DM nightmare...no not really a nightmare. It is the expected result of running at level 11 or 12 and allowing players to run ridiculous OP subclasses and builds.
You are going to have to ignore the CR values of monsters, and give your monsters many many Legendary Resistances and Actions. You will have to give them minions. And you will have to make sure the BBEG is airborne and launching death from over 100 feet above the group. Take an ancient Red Dragon, then tack on all the abilities and spells of say a 15th level Divine Soul Sorcerer, and that would be a start.And EVEN THEN, thanks to Sentinel, if a PC gets close, the fight is over.
Although I do agree that Eldritch Smite is OP, and even ridiculous, it really just needs a few minor tweaks to be brought back into the realm of reasonable. It should be reduced to large or smaller, or even medium or smaller and require the target to make a save, on a fail they get knocked prone.
Anyone who has DMed, 5e knowns that the CR system is a baseline and is not a hard line, also there is a lot of miss use, misunderstanding of how to balance an adventuring day, with encounters, short rest, long rest, ect, using the CR system. I usually find that people who are having issues with the CR system aren't using it correctly. and Yes, we have to remember it is based on 5e (kinder, gentler, DnD) not OD&D or 2e.
"Give your monsters many many Legendary Resistances and Actions. You will have to give them minions. And you will have to make sure the BBEG is airborne and launching death from over 100 feet above the group. Take an ancient Red Dragon, then tack on all the abilities and spells of say a 15th level Divine Soul Sorcerer, and that would be a start." that would not be a start that would be a TPK, and intentional one. That Sorcerer build alone if played right could TPK a level 11 party in like maybe one round, two at the most. That's without giving it legendary actions, legendary resistances, a breath weapon that does 26d6 fire damage, and 546 hit points.
While it's true that a significant part of the problem with the CR system is that people don't really get the six encounter day, the fact that the CR system is based around a play style that no-one actually wants to use is itself a problem. However, independent of that, the CR system has a math problem: PC power scales up faster than monster power. A typical CR 20 has about 3.5x the damage and defense of a CR 5; a typical level 20 has something like 3x the offense and 8x the defense of a level 5, not to mention a likely boatload of weird powers that can end encounters in a way that doesn't directly interact with monster power. As a result, while a reasonable encounter at level 5 is something like 50% of the daily budget, a reasonable encounter at level 20 is more like 100%.
Firstly, DnD 5e is not design for players to die both in that it isn't very likely and that it isn't made fun. It doesn't have the rogue like elements of earlier editions where you have random characters and an xp scale that encourages catch up. I wouldn't recommend trying to kill players for that reason which means your tension needs to come from else where. Set goals that they can fail in ways other than the players dieing. For example give them rescue or capture alive missions. Even rats can be challenging if you are trying to stop them killing commoners as opposed to demigod players. You can go full plague tale on them.
Secondly If they are performing as a higher level it only makes sense that they will be asked to and may try more challenging tasks. Consider treating them as a party a couple of levels higher if you are going to make the challenge putting the players at risk.
Thirdly there are ways to put players at risk other than just monsters and you should try a variety of different things to keep it interest and avoid turning it into a tit v.s tat of optimizing. Use things like:
Finally there are elements of encounter design that can reduce the effects of certain strategies.
For sure! I don't know any DMs that run 6 CR based/combat encounters in an adventuring day, and I don't know any players that would want it that way. But as a result we end up with people running one or two easy encounters, or they skip the easy and do one or two medium encounter, then one hard or deadly, and let the players take short or long rest in-between, and wonder why the party runs all over everything.
As far as the math problem. I think, although there may very well be some mathematical issues, I think it is just as much a perception issue. Yes players level/scale faster, but there is zero correlation between player level and CR as far as as level 5 PC has nothing to do with a CR 5 monster, and no correlation between CR and PC level should be made. Also saying that CR20 should be 15x more than a CR5 was is not how the CR system works and it can't work that way. If that is how the CR system worked all CR10 enemies would be 100% identical with pasted on names and themes. In a system which worked that way a CR15 Dragon and a CR 15 Archfiend would have the exact same health, damage output, and abilities, and that would be boring. Instead the CR system is a baseline system which takes averages of monsters and categorizes them for the ease of selecting what the average difficulty will be against an average party with X number of members. That's how you end up with two CR12 monsters with sometimes extremely varied hp, damage output, and abilities. A monster with low HP and high damage might be the same CR as a monster with high hp and lower Damage, depending on the other ability's of the monster.
The concept behind CR is that a CR x monster is a medium encounter for a level x party of 4 -- level 5 party gets 1xCR 5, level 20 party gets 1xCR 20 (the math is slightly erratic, but it's mostly true), and if you look at the standards for CR x monsters, it turns out that things like AC are tuned to match the expected capabilities of level x PCs.
The CR system is designed around expected damage output and hit points. There are some anomalous monsters, but for the most part you can assert "A CR x monster has (x+1)*15 hp and does (x+1) damage per round, with an xp value of (x+1)*(x+1)*50, possibly with some tradeoff of being above-CR on one measure and below on the other" and be ballpark correct up through CR 20.
That may very well be the intended design concept( I have no idea), but if you go into encounter builder right now and set up a 4 member level 5 party and then filter so you only see level 5 monsters, you get easy encounters, not medium, With a 4 member level 5 party CR6-CR7 is medium, CR8 is hard, and CR9-whatever is deadly, there should be a feature that says impossible so you know....but is doesn't, I'm guessing anything above a CR11 or CR12 would be impossible for a 4 member level 5 part. So if you filter the encounter builder for CR5 versus a level 5 party of for 4, and you look at the AC for the first 25 monsters the AC ranges from 9 to 19 with no meaningful average. In the same scenario you get hit points from 67 to 135, that's a 68 point swing. With HP you at least have an average of 90-114 ish 10 out of 25 were in this range. This whole concept completely breaks down at CR1 CR2 a party of 4 level 1 Anythings against some of the CR-1 monsters PC's are gonna die. Same goes for CR2 at these CR levels, when its already hard not to accidently TPK, I would not trust the CR system at all CR1-3.
The expected damage out put and hit points doesn't really gel either. With the HP 90ish is the low end of the average which adds up with your math (5+1)*15 =90 hp. But in the same 25 CR5 sample the average was over 15 points of damage, that was assuming they hit 50% of the time. (x+1) damage in this sample should be (5+1) =6 damage per round. The average was more then double that (in the small sample I did), but there was not a single monster in the sample that only did (x+1) damage per round.
Realistically none of this matters because if you are running a party of 4 level 5 Hill Dwarf barbarians. All of these CR 5 monsters would be toast. Conversely a party of 4 level 5 Wizards doing average damage of their highest level spell slots might take out anything in the 90-114 hit point category before any of them die, but all the wizards would need to go first, and the CR5 monster would need to fail all of its saves. In this scenario there would be a dead wizard by the CR5 monster's second turn, making this easy encounter a deadly one.
All of that was my longwinded way of saying the CR system is still just a base line. Its a good way to gage if your homebrew monsters are in the ball park for your party. But ultimately you have to know your party and what they are capable of. In my experiences the CR system usually errors in favor of going easy on the party at least CR level 5-15ish, before that it seems like it errors on the side of killing the PCs.
Sorry, typo. Supposed to be (x+1)x5.
lol, that makes way more sense!
If you ignore a few of the bad assumptions Cr seems to make it does describe monsters fairly accurately, the issue is that there is a huge range in the power of player builds. This is partly due to what 5e considers optional rules because encounter balance doesn't take into account things like feats, multiclassing and magic items which many parties treat as standard. Its also out dated because the newer classes and subclasses tend to be stronger than many of the launch options.
If you really have no clue of where the party should be then you should be able to match it to the Create a monster table. Average the hit chance and ac and match it to the table, for health and damage compare the sum of the parties health and damage to the table. You can then treat the party as that level and CR should work a little better. However I think you inevitably reach a level where you need to be more reactive and just adjust the difficulty up and down based on how encounters go.
I also think that by the time you reach level 10 encounters need to start being more than just combat. Legendary monster like vampires become easily defeatable so unless you are going to throw demigods at them on the regular the goals need to be more than just simply killing them.