Also perhaps that formula is meant specifically for homebrewing critters and serves as a guesstimate framework rather than a complete thing.
Not really "perhaps". I'm 100% certain WotC devs use a different formula, which logically means the one in the DMG is only meant for DMs creating their own monsters. I'm equally certain that while not perfect WotC's in-house formula is better than the one in the DMG, it's also a lot more complex and that that is the reason they deemed it a bad idea to put it in the DMG.
Rollback Post to RevisionRollBack
Want to start playing but don't have anyone to play with? You can try these options: [link].
Also perhaps that formula is meant specifically for homebrewing critters and serves as a guesstimate framework rather than a complete thing.
Not really "perhaps". I'm 100% certain WotC devs use a different formula, which logically means the one in the DMG is only meant for DMs creating their own monsters. I'm equally certain that while not perfect WotC's in-house formula is better than the one in the DMG, it's also a lot more complex and that that is the reason they deemed it a bad idea to put it in the DMG.
From what I recall (and I can't remember where I'm recalling this from, so I apologise), they use pretty much the same process albeit with some finer granularity on the numbers. The 'secret sauce' is playtesting; they'll run the monsters though the ringer internally and make tweaks. This is why you see some monsters with seemingly wrong values for things.
If I can find where I read/heard this, I'll edit it to this post.
Also perhaps that formula is meant specifically for homebrewing critters and serves as a guesstimate framework rather than a complete thing.
Not really "perhaps". I'm 100% certain WotC devs use a different formula.
Someone did some analysis here. It isn't all that different offensively (the DMG formula works out to +1 to hit = +3.6 dpr and you need 7.2 effective dpr (6 damage, 1/3 to hit) for +1 CR; the actual formula is more like 6.8 (5 damage, 1/2 to hit) but the DMG gives low CR monsters rather high hit points.
You want a benchmark, that's what you get. There are literally hundreds of possible variables involved in a high level combat encounter and that's not counting dice rolls. It ain't going to get more accurate. This is especially true with a single monster without legendary and/or lair actions because anything there boils down to how much damage it can output before a "typical" party hits it enough to kill it and that is what CR is based on. It's the amount of damage it can do in it's "optimal action path" th.at Crawford talks about in that video. It's what he flat out says they're trying to dumb the stat block down to so DMs who can't be bothered to actually think for themselves don't get distracted by other things in the entry.
In my example with the storm giant the monster for that "medium" encounter probably could take out the Wizard, or at least down them, by focusing on them before the rest of the party kills it. That's potentially "deadly." That's also an encounter that'll probably be over in two or maybe three rounds. And you're going to have a pissed off Wizard player for being targeted by the big stick of a monster that's using it's "optimal damage path" to neutralize the biggest threat to it and it's low Dex save so it might survive long enough to also injure another PC before it dies.
You want longer encounters that are more nuanced than a microcosm of an ICBM exchange, use the tables and build an encounter of similar difficulty with a larger number of weaker creatures. But then if you only use the dumbed down "optimal path" for each one fighting as an individual you're just going to have a bunch of weenies getting taken out by a single AoE spell or such because they're all bum-rushing straight in. By adding more pieces to the board you have to thinkto figure out how to make them effective enough to not just die like a swarm of gnats flying into a bug zapper represented by the party. And the stat blocks don't include things like "when part of a larger group, this creature should attack from range, separate from other range attackers, and attempt to aid melee focused allies" because that's common sense and Crawford said they're assuming DMs have at least some of that. You aren't going to find common sense in a stat block, just numbers that a moderately intelligent fourteen year old should be able to figure out on their own with a few minutes to think about it.
The CR system also depends on a "typical" party that is balanced as the game is designed to be. That means no homebrew abilities or classes that are more often than not based on what the players and/or DM thinks is cool with no informed consideration to actual game balance. Or, as someone already mentioned, using "optional rules" like flanking and feats that the base game design does not factor (or at least does not factor heavily) because those are extra variables that make the entire convoluted tangle of the basic rules even more complex and unpredictable. You can get balanced encounters with those optional rules but you're going to have to actually think about things.
CR simply cannot be more "accurate" than it already is in any substantial and noticeable degree. It's a complex game and any stability derived from increasing margin of error for a random crit because of larger HP pools is offset by the exponentially greater other variables that come from additional character options as levels go up. It is impossible for it to be simplified to a degree that CR is anything nearer to universally "accurate" as it is, and even then it's intended for use over multiple encounters per long restinstead of the "one monster one fight" nukefest that those complaining about this situation seem to expect can be simplified to be one size fits all satisfying battles.
TLDR: To reference a popular meme template, "Balanced combat with a single monster? The action economy is in shambles." No, that doesn't really make sense. Neither do the expectations of anybody that wants CR to be "fixed" so they don't have to think about their encounters.
Zero, of course :). The DMG doesn't count status effects towards CR at all. Which is why a pixie is CR 1/4; despite a vast number of obnoxious abilities, its only ability to actually do damage is Phantasmal Force.
Heh. You know that, I know that, but a lot of people apparently don't know that. All the people demanding absolutely pitch-perfect, one hundred percent consistent mathemagical algorithms that perfectly encapsulate a given critter's threat level are also people who, I presume, would never use pixies at their table because pixies are useless* in open combat. Meanwhile some of us still pine for the ability to play the durned things as PCs and know full well the value of properly timed fey mischief. Not all encounters involve axefacing, and as Pang stated/logic would dictate, extensive playtesting will show occasional edge cases that need adjustment where the algorithm fails.
All of which is among the many reasons I just kinda roll my eyes at people who try and demand a perfect, always-accurate algorithm for mathematically determining encounter balance. It's almost like even oversimplified 5e is a giant pile of semi-randomized bits and pieces, wild and outlandish powers, abilities, and side effects, that tends to produce emergent results that fall beyond the narrow scope of Encounter Damage Math regularly.
*"useless" in the sense that the pixie cannot deal damage save, as mentioned, with Phantasmal Force or "indirectly" via Polymorph. Anyone who believes pixies are no threat because they cannot deal HP damage probably doesn't deserve their character sheet/DM screen.
People want Challenge Rating to be absolute and infallible - input Party X, perform mathematical operation Y, get result Z, your Perfectly Balanced Encounter. People - rightly - naysay that idea because it's hogwash. At low levels, one single crit can throw off an entire encounter. At high levels, players have enough options that encounter calculus becomes more trouble than it's worth. Rigorous, exacting mechanical balance, tuned to within an inch of its life and predictable down to the last die roll, is not only difficult but also unnecessary, especially when the system doesn't work as much due to DMs being as bad at combat as many of their players are. Not every DM is a tactical mastermind; the whole monster statblock shifts coming in M3 to emphasize a monster's primary actions is a measure to help DMs stop self-sabotaging themselves with more complex monsters.
Ultimately that's why people keep carping on CR. It's like the infinite *****ing about matchmaking algorithms in PvP video games - people want the system to do all the work of making matches/encounters for them and they refuse to tolerate "machine code can't account for the human factor in your game" as an answer.
What you say is not at all true for most of us. We know encounters may hinge on crits etc. We know we have to adjust for player skill, composition, size, rest frequency, etc. But we want a solid BENCHMARK from which to adjust encounter difficulty. We want and expect CR to follow the parameters indicated in the DMG. We want and expect that the monsters' CRs be assigned based on those parameters. They aren't, because the designers did a poor job of accurately assigning those CRs. This is a well documented fact.
Isolating variables is fundamental to logic, math, and the scientific method. The CR rating is one variable that we expect the designers to have done accurately. They don't need to account for party size, composition, skill level, or rest frequency. Player skill is on the DM to adjust for, as is party composition. Party size has guidelines for adjustment given in the DMG. The only variable that then remains is luck. By limiting the variables in this way to 1, we decrease the standard deviation of encounter balance.
In other words, the DMG gives me a CR x monster and tells me how to adjust for party size. Then I adjust up or down x levels based on party skill and rest frequency. Then I look at the party composition and move up or down accordingly. Then it all comes down to luck - sure the monster may crit 4 times and make the encounter unbalanced, but it wont happen often, and the end product is much closer to balanced. But because there are so many inaccuracies in monster CR ratings assigned (and to be clear, I mean inaccurate within their own guidelines, not mine), everything is thrown off x% of the time.
The argument is about expecting a CR rating to alleviate any need for adjustment, and/or to be a "one stop shop" for encounter building, is a straw man at this point. While a few peeps might think that way, for the vast majority of us complaining about CR, its about giving us a solid and consistent benchmark to adjust from. Its about their failure to use their own system to calculate said CRs consistently.
You want a consistent benchmark for encounter building? Go old school... Total the party Hit Dice and double it. That is the total number of hit dice the foes need to make the encounter challenging. Ignore CR it has never worked (not even in 3.x where it was first introduced).
You want a consistent benchmark for encounter building? Go old school... Total the party Hit Dice and double it. That is the total number of hit dice the foes need to make the encounter challenging. Ignore CR it has never worked (not even in 3.x where it was first introduced).
Ever since it was introduced, CR suffered from the problem that they were balancing for a way people didn't actually play -- the assumption is that people will plow through a half dozen encounters before taking a long rest, and most players and DMs don't particularly want to do that. Standard Monster of the Week encounter design, which is what most DMs will default to and most players will expect, involves 1-3 encounters: one encounter that introduces the threat, one midpoint encounter, one boss fight, where the first and second encounters might not actually happen.
You want a consistent benchmark for encounter building? Go old school... Total the party Hit Dice and double it. That is the total number of hit dice the foes need to make the encounter challenging. Ignore CR it has never worked (not even in 3.x where it was first introduced).
Consistent is one thing. Consistently bad is quite another.
Rollback Post to RevisionRollBack
Want to start playing but don't have anyone to play with? You can try these options: [link].
I think the primary problem is that the encounter building guidelines make people think that a fight that consumes 1/4 of your daily resources is Hard.
That was, I believe, the goal of CR equivalent to a four member party in previous editions (pretty sure I remember something to that effect from 3.x). There is nothing that alludes to this in the encounter difficulty descriptions in the 5e DMG, which are separated by difficulty, not CR (even though the difficulty is based on "experience thresholds" which are tied to CR), and the "definitions" of those difficulty levels are general and vague. Nowhere in that entire section does anything promise a fixed "difficulty" based on CR. I seriously doubt that most of the people complaining about how imprecise CR is have actually fully read that section.
And Yurei makes excellent points to accentuate how utterly impossible it is to "accurately" categorize any given monster's threat level even before factoring in the wildly varied abilities of any given party. CR provides a general ballpark figure and then the DM has to think, preferably applying knowledge of how combat works and how the various abilities of monsters, PCs, NPCs, etc involved in order to balance the encounter.
I think the primary problem is that the encounter building guidelines make people think that a fight that consumes 1/4 of your daily resources is Hard.
That was, I believe, the goal of CR equivalent to a four member party in previous editions (pretty sure I remember something to that effect from 3.x). There is nothing that alludes to this in the encounter difficulty descriptions in the 5e DMG..
Well, there is in the overall section, though it's down in the Adventuring Day: "Assuming typical adventuring conditions and average luck, most adventuring parties can handle about six to eight medium or hard encounters in a day." It's also implicit in the daily budget.
Doubling all encounter budgets (without changing daily budget), so PCs are now expected to handle 3-4 medium encounters per day, produces results that are a much better match to how games are actually played.
"Six to eight" times 1/4 would consume 150-200% of the party's resources, exhausting all spells and other consumable abilities and killing them all up to twice. This clear inconsistency is likely due to nothing in that statement referring to the number "1/4."
"Six to eight" times 1/4 would consume 150-200% of the party's resources, exhausting all spells and other consumable abilities and killing them all up to twice. This clear inconsistency is likely due to nothing in that statement referring to the number "1/4."
It's actually 6-8x the medium encounter threshold. The Hard encounter threshold is typically about 1/4 of the daily budget.
Modern WotC cant even balance simple spells or Subclasses, and you expect them to be able to fix CR?
The people currently in charge of D&D have no clue what they are doing.
The most balanced version of D&D was 4E, and that edition was disliked by many specifically because of choices made to facilitate that balance. Third edition was an all-you-can-eat buffet for ruleslawyers looking for exploitable loopholes and builds that were optimized just so. TSR's D&D thought it a good idea to lock mechanically stronger classes behind needing to roll higher stats, so you essentially either won twice or lost twice, and their idea of wizards as a balanced class was that being gods at high level was evened out by being runty wimps at low levels.
Balance is what you make of it. And that goes for pretty much any TTRPG I've ever played. Expecting balance straight out of a book is beyond naive.
Rollback Post to RevisionRollBack
Want to start playing but don't have anyone to play with? You can try these options: [link].
OK I will say this again, 20 years experiance in TTRPG's and 5 years DnD 5e experiance (my first DnD edition I ever played/DM'd). I have yet to find a TTRPG system that gets an equivalent CR system right, and I have played many many many different systems. Many TTRPG's don't even bother trying to do one so the fact DnD does can be considered either a good thing in terms of having a starting point, or a bad thing because it isn't great.
Personally I don't use it and I don't see the need for it really.
people who say a balanced cr system doesn't exist I point you to pathfinder 2e as someone who played 5e from its release until 2020 then switched yeah its actually balanced if the rules say something is severe it actually is I highly suggest looking into it. & no it is not a 1st edition so no insane modifier stacking. it also has way more classes dm material rules for dms to use better pc customization & better supplement books and adventure modules next month paizo is releasing book of the dead with anti undead options for players playable undead for pcs full vampirism rules for playing a lich goul or ghost a skeleton race & tons of new undead creatures & necromancy spells
cr is one of the constant complaints I see about 5e pathfinder2e providers cr that actually works!!!
You want a consistent benchmark for encounter building? Go old school... Total the party Hit Dice and double it. That is the total number of hit dice the foes need to make the encounter challenging. Ignore CR it has never worked (not even in 3.x where it was first introduced).
Consistent is one thing. Consistently bad is quite another.
The old school method hasnt failed me yet. It gives me a ball park to start with and then adjust from there based on party composition.
I just had my level 15 characters make short work of a CR 19 creature. It has been clear to me for a long time now that CR ratings are useless. Is WOTC planning to fix this with 5.5e? It really is becoming a sore point seeing my players have no problem during combat.
Using the new system from XGtE, a single monster that has a CR two to four higher than the average party level would be a balenced encounter.
Were any of your players min-maxing/optimizing? What was the party's composition?
A similar thing happens in a group I am playing in. We mauled two bone devils as five level-10 characters. That would be a deadly encounter, but we torn those fiends apart like if it was an easy encounter. Then again, my character had firearms and Sharpshooter, while another character was a Hexblade Warlock with Great Weapon Master. We also had a shield guardian on our side, belonging to our Grave Cleric.
If you feel that your encounters are too easy for your players, there are two things you can do: use higher CR monsters, or buff the hit points of your current monsters (maybe even add a few extra damage dice). That's what my DM does, because our party is kind of OP. He only buffs the hit points, but doesn't change much else about the monster (for simplicity and adherence to RAW).
Rollback Post to RevisionRollBack
Brains over brawn? Mind over matter? These canny warriors rightly answer, "Why not both?" - Tasha
Not really "perhaps". I'm 100% certain WotC devs use a different formula, which logically means the one in the DMG is only meant for DMs creating their own monsters. I'm equally certain that while not perfect WotC's in-house formula is better than the one in the DMG, it's also a lot more complex and that that is the reason they deemed it a bad idea to put it in the DMG.
Want to start playing but don't have anyone to play with? You can try these options: [link].
From what I recall (and I can't remember where I'm recalling this from, so I apologise), they use pretty much the same process albeit with some finer granularity on the numbers. The 'secret sauce' is playtesting; they'll run the monsters though the ringer internally and make tweaks. This is why you see some monsters with seemingly wrong values for things.
If I can find where I read/heard this, I'll edit it to this post.
Find my D&D Beyond articles here
Someone did some analysis here. It isn't all that different offensively (the DMG formula works out to +1 to hit = +3.6 dpr and you need 7.2 effective dpr (6 damage, 1/3 to hit) for +1 CR; the actual formula is more like 6.8 (5 damage, 1/2 to hit) but the DMG gives low CR monsters rather high hit points.
You want a benchmark, that's what you get. There are literally hundreds of possible variables involved in a high level combat encounter and that's not counting dice rolls. It ain't going to get more accurate. This is especially true with a single monster without legendary and/or lair actions because anything there boils down to how much damage it can output before a "typical" party hits it enough to kill it and that is what CR is based on. It's the amount of damage it can do in it's "optimal action path" th.at Crawford talks about in that video. It's what he flat out says they're trying to dumb the stat block down to so DMs who can't be bothered to actually think for themselves don't get distracted by other things in the entry.
In my example with the storm giant the monster for that "medium" encounter probably could take out the Wizard, or at least down them, by focusing on them before the rest of the party kills it. That's potentially "deadly." That's also an encounter that'll probably be over in two or maybe three rounds. And you're going to have a pissed off Wizard player for being targeted by the big stick of a monster that's using it's "optimal damage path" to neutralize the biggest threat to it and it's low Dex save so it might survive long enough to also injure another PC before it dies.
You want longer encounters that are more nuanced than a microcosm of an ICBM exchange, use the tables and build an encounter of similar difficulty with a larger number of weaker creatures. But then if you only use the dumbed down "optimal path" for each one fighting as an individual you're just going to have a bunch of weenies getting taken out by a single AoE spell or such because they're all bum-rushing straight in. By adding more pieces to the board you have to think to figure out how to make them effective enough to not just die like a swarm of gnats flying into a bug zapper represented by the party. And the stat blocks don't include things like "when part of a larger group, this creature should attack from range, separate from other range attackers, and attempt to aid melee focused allies" because that's common sense and Crawford said they're assuming DMs have at least some of that. You aren't going to find common sense in a stat block, just numbers that a moderately intelligent fourteen year old should be able to figure out on their own with a few minutes to think about it.
The CR system also depends on a "typical" party that is balanced as the game is designed to be. That means no homebrew abilities or classes that are more often than not based on what the players and/or DM thinks is cool with no informed consideration to actual game balance. Or, as someone already mentioned, using "optional rules" like flanking and feats that the base game design does not factor (or at least does not factor heavily) because those are extra variables that make the entire convoluted tangle of the basic rules even more complex and unpredictable. You can get balanced encounters with those optional rules but you're going to have to actually think about things.
CR simply cannot be more "accurate" than it already is in any substantial and noticeable degree. It's a complex game and any stability derived from increasing margin of error for a random crit because of larger HP pools is offset by the exponentially greater other variables that come from additional character options as levels go up. It is impossible for it to be simplified to a degree that CR is anything nearer to universally "accurate" as it is, and even then it's intended for use over multiple encounters per long restinstead of the "one monster one fight" nukefest that those complaining about this situation seem to expect can be simplified to be one size fits all satisfying battles.
TLDR: To reference a popular meme template, "Balanced combat with a single monster? The action economy is in shambles." No, that doesn't really make sense. Neither do the expectations of anybody that wants CR to be "fixed" so they don't have to think about their encounters.
Heh. You know that, I know that, but a lot of people apparently don't know that. All the people demanding absolutely pitch-perfect, one hundred percent consistent mathemagical algorithms that perfectly encapsulate a given critter's threat level are also people who, I presume, would never use pixies at their table because pixies are useless* in open combat. Meanwhile some of us still pine for the ability to play the durned things as PCs and know full well the value of properly timed fey mischief. Not all encounters involve axefacing, and as Pang stated/logic would dictate, extensive playtesting will show occasional edge cases that need adjustment where the algorithm fails.
All of which is among the many reasons I just kinda roll my eyes at people who try and demand a perfect, always-accurate algorithm for mathematically determining encounter balance. It's almost like even oversimplified 5e is a giant pile of semi-randomized bits and pieces, wild and outlandish powers, abilities, and side effects, that tends to produce emergent results that fall beyond the narrow scope of Encounter Damage Math regularly.
*"useless" in the sense that the pixie cannot deal damage save, as mentioned, with Phantasmal Force or "indirectly" via Polymorph. Anyone who believes pixies are no threat because they cannot deal HP damage probably doesn't deserve their character sheet/DM screen.
Please do not contact or message me.
You want a consistent benchmark for encounter building? Go old school... Total the party Hit Dice and double it. That is the total number of hit dice the foes need to make the encounter challenging. Ignore CR it has never worked (not even in 3.x where it was first introduced).
Ever since it was introduced, CR suffered from the problem that they were balancing for a way people didn't actually play -- the assumption is that people will plow through a half dozen encounters before taking a long rest, and most players and DMs don't particularly want to do that. Standard Monster of the Week encounter design, which is what most DMs will default to and most players will expect, involves 1-3 encounters: one encounter that introduces the threat, one midpoint encounter, one boss fight, where the first and second encounters might not actually happen.
Consistent is one thing. Consistently bad is quite another.
Want to start playing but don't have anyone to play with? You can try these options: [link].
That was, I believe, the goal of CR equivalent to a four member party in previous editions (pretty sure I remember something to that effect from 3.x). There is nothing that alludes to this in the encounter difficulty descriptions in the 5e DMG, which are separated by difficulty, not CR (even though the difficulty is based on "experience thresholds" which are tied to CR), and the "definitions" of those difficulty levels are general and vague. Nowhere in that entire section does anything promise a fixed "difficulty" based on CR. I seriously doubt that most of the people complaining about how imprecise CR is have actually fully read that section.
And Yurei makes excellent points to accentuate how utterly impossible it is to "accurately" categorize any given monster's threat level even before factoring in the wildly varied abilities of any given party. CR provides a general ballpark figure and then the DM has to think, preferably applying knowledge of how combat works and how the various abilities of monsters, PCs, NPCs, etc involved in order to balance the encounter.
Well, there is in the overall section, though it's down in the Adventuring Day: "Assuming typical adventuring conditions and average luck, most adventuring parties can handle about six to eight medium or hard encounters in a day." It's also implicit in the daily budget.
Doubling all encounter budgets (without changing daily budget), so PCs are now expected to handle 3-4 medium encounters per day, produces results that are a much better match to how games are actually played.
"Six to eight" times 1/4 would consume 150-200% of the party's resources, exhausting all spells and other consumable abilities and killing them all up to twice. This clear inconsistency is likely due to nothing in that statement referring to the number "1/4."
It's actually 6-8x the medium encounter threshold. The Hard encounter threshold is typically about 1/4 of the daily budget.
In the Encounter builder a party of 4 x level 15 against a CR 19 is simply a hard encounter.
And is just over 30% of the daily budget. In my revised difficulty it would listed as Easy (though very close to Moderate).
Modern WotC cant even balance simple spells or Subclasses, and you expect them to be able to fix CR?
The people currently in charge of D&D have no clue what they are doing.
The most balanced version of D&D was 4E, and that edition was disliked by many specifically because of choices made to facilitate that balance. Third edition was an all-you-can-eat buffet for ruleslawyers looking for exploitable loopholes and builds that were optimized just so. TSR's D&D thought it a good idea to lock mechanically stronger classes behind needing to roll higher stats, so you essentially either won twice or lost twice, and their idea of wizards as a balanced class was that being gods at high level was evened out by being runty wimps at low levels.
Balance is what you make of it. And that goes for pretty much any TTRPG I've ever played. Expecting balance straight out of a book is beyond naive.
Want to start playing but don't have anyone to play with? You can try these options: [link].
OK I will say this again, 20 years experiance in TTRPG's and 5 years DnD 5e experiance (my first DnD edition I ever played/DM'd). I have yet to find a TTRPG system that gets an equivalent CR system right, and I have played many many many different systems. Many TTRPG's don't even bother trying to do one so the fact DnD does can be considered either a good thing in terms of having a starting point, or a bad thing because it isn't great.
Personally I don't use it and I don't see the need for it really.
people who say a balanced cr system doesn't exist I point you to pathfinder 2e as someone who played 5e from its release until 2020 then switched yeah its actually balanced if the rules say something is severe it actually is I highly suggest looking into it. & no it is not a 1st edition so no insane modifier stacking. it also has way more classes dm material rules for dms to use better pc customization & better supplement books and adventure modules next month paizo is releasing book of the dead with anti undead options for players playable undead for pcs full vampirism rules for playing a lich goul or ghost a skeleton race & tons of new undead creatures & necromancy spells
cr is one of the constant complaints I see about 5e pathfinder2e providers cr that actually works!!!
The old school method hasnt failed me yet. It gives me a ball park to start with and then adjust from there based on party composition.
Using the new system from XGtE, a single monster that has a CR two to four higher than the average party level would be a balenced encounter.
Were any of your players min-maxing/optimizing? What was the party's composition?
A similar thing happens in a group I am playing in. We mauled two bone devils as five level-10 characters. That would be a deadly encounter, but we torn those fiends apart like if it was an easy encounter. Then again, my character had firearms and Sharpshooter, while another character was a Hexblade Warlock with Great Weapon Master. We also had a shield guardian on our side, belonging to our Grave Cleric.
If you feel that your encounters are too easy for your players, there are two things you can do: use higher CR monsters, or buff the hit points of your current monsters (maybe even add a few extra damage dice). That's what my DM does, because our party is kind of OP. He only buffs the hit points, but doesn't change much else about the monster (for simplicity and adherence to RAW).
Brains over brawn? Mind over matter? These canny warriors rightly answer, "Why not both?" - Tasha
My Homebrews: Monsters, Magic Items, Spells, Races
Rhulg- Hobgoblin Gunsmith