D&D 5E Encounter Building math: or I killed 2 PCs last night.

[keterys]

I thought CR and XP _were_ divorced? Ie, CR is more of a warning system for when something is reasonable to throw at a party, and XP is how you actually balance something.

It may be that a few monsters are incorrectly XP-ed (or CR-ed), but the base concept seems fine. It is probably reasonable to identify monsters whose XP values are wrong.

It's also probably reasonable for monsters with strong synergies to have different XP values depending on the number encountered at once. A hobgoblin or kobold alone is a far cry from a band of 10 of them, for example.

 

[Savage Wombat]

I'm not the person you're quoting (obviously), but I think we would have been better served by divorcing XP from CR. That would allow more granularity with encounter building while keeping CR as a general guideline of toughness.

I've been saying this since we saw the first monster guidelines. I wonder if some math-oriented ENWorlder with time on his hands wants to take a crack at this?

 

[wedgeski]

I thought CR and XP _were_ divorced? Ie, CR is more of a warning system for when something is reasonable to throw at a party, and XP is how you actually balance something.

Yup. The designers have identified the problem that encounter building is at least a two-dimensional problem, and designed accordingly.

It's also probably reasonable for monsters with strong synergies to have different XP values depending on the number encountered at once. A hobgoblin or kobold alone is a far cry from a band of 10 of them, for example.

There has been a specific effort to address the multiple enemy problem (see example near OP), but synergies are next to impossible to solve mathematically. The DM has to take a role at some point!

 

[drjones]

Yup. The designers have identified the problem that encounter building is at least a two-dimensional problem, and designed accordingly.

I think the issue there is that the draft DM basic rules maps CR to Xp directly, meaning they are direct replacements for each other. But I believe that table is just supposed to be a guideline (and I don't think every monster in the MM follows the table, which is good) so if you make a new monster and determine that a level 2 PC would have the resources to deal with it but it is not as dangerous as other CR2 monsters you would make it CR2 but give less XP than the table would infer. At least I hope that's how it works, these rules are still in the oven.

 

[Tormyr]

Is the level of granularity really necessary? Does it really make a difference?
CR1 200XP
CR2 450XP
CR3 700XP

So if a creature had an XP range, say halfway between the average CRs on the low and high side, does it make a difference? In that scenario, a CR2 creature could range from 325XP to 575XP. Does it really make a difference to have an intellect devourer have 575XP instead of 450XP?

As it stands in the current encounter building guidelines, 1 intellect devourer (450XP) is a moderate challenge for a 4 person level 2 party, and 2 (1,350XP) are beyond deadly. Any other CR2 creature from Ogres to Saber-Tooth Tigers are the same. 1 is moderate, 2 is deadly.

If the intellect devourer was at 575XP, It would be hard for the level 2 party of 4 and 2 of them would still be deadly, but is an intellect devourer really hard for the level 2 party? It has lousy AC and low hit points for a CR2 creature, on par with a CR2 wizard. The party of 4 is going to take it down in 1 round, and the creature will be fortunate to get 1 shot off. The claws are not very powerful for a CR2 creature. Devour Intellect requires a failure on a DC12 saving throw or take (11) 2d10 psychic damage and roll 3d6 to meet or exceed the target's intelligence or the target's int goes to 0. Most of the time, the level 2 wizard (10hp) has to worry about being one-shotted by a CR2 ogre. This time the fighter has to worry about being one-shotted.

As it stands 2 CR2s (1350XP) cross the moderate threshold going from a level 4 (1000XP) to a level 5 (2000XP) party. at 575XP each, 2 intellect devourers (1725XP) would be cross the threshold from hard to moderate as the party went from level 4 to level 5. So what is the difference? The intellect devourer is not really more deadly than any other CR2 creature, so why complicate encounter building? To steal and modify a phrase from something else entirely, the encounter building guidelines should be, "As simple as possible; as complex as necessary." The simplicity comes in all creatures of a CR having the same XP. The complexity comes from the difficulty not scaling linearly with the number of creatures.

Now with calculators in phones and spreadsheets in laptops, calculating XP budgets when monsters of the same CR have different XP is only mildly more work, but I would argue that, while its uber attack is more of a pain to deal with than other CR2 creatures, the intellect devourer is not really more deadly than other CR2 creatures and does not really need a modest bump in XP over other CR2 creatures.

Obviously, the intellect devourer is just one example, but I think this example would mostly hold true for just about any creature compared to other creatures of its CR with just the numbers changing.

 

[keterys]

There has been a specific effort to address the multiple enemy problem (see example near OP), but synergies are next to impossible to solve mathematically. The DM has to take a role at some point!

Except their "just multiply the whole thing by a bunch as you add more" theory is really awful. It's the opposite of what you should do for monsters without synergies, especially melee-centric ones who will have difficulty bringing their might to bear and are easily taken down by AE, and the math/break points on it are silly.

On the other hand, some monsters have extraordinary synergies that could just be called out. Intellect Devourer could be double xp when grouped with a monster that incapacitates (stuns), triple xp when grouped with other intellect devourers. Now DMs won't be surprised when one intellect devourer sets up the other to automatically devour the brain of a PC. And it's not like the math is fiddlier than what we have (which is pretty awful). Hobgoblins have synergy with, well, everything, so are pretty crappy XP on their own, but great XP as soon as you add any number of melee-capable allies.

P.S. The really sad thing about the intellect devourer example is that it doesn't really get much easier as you level. It's not like your save bonus goes up much, so running into 10-20 of them at high level? Better not be surprised, win initiative, and hope you can kill almost all of them before they go.

 

[Tormyr]

Except their "just multiply the whole thing by a bunch as you add more" theory is really awful. It's the opposite of what you should do for monsters without synergies, especially melee-centric ones who will have difficulty bringing their might to bear and are easily taken down by AE, and the math/break points on it are silly.

On the other hand, some monsters have extraordinary synergies that could just be called out. Intellect Devourer could be double xp when grouped with a monster that incapacitates (stuns), triple xp when grouped with other intellect devourers. Now DMs won't be surprised when one intellect devourer sets up the other to automatically devour the brain of a PC. And it's not like the math is fiddlier than what we have (which is pretty awful). Hobgoblins have synergy with, well, everything, so are pretty crappy XP on their own, but great XP as soon as you add any number of melee-capable allies.

P.S. The really sad thing about the intellect devourer example is that it doesn't really get much easier as you level. It's not like your save bonus goes up much, so running into 10-20 of them at high level? Better not be surprised, win initiative, and hope you can kill almost all of them before they go.

Huh, I just don't see the encounter building guidelines that negatively. For one thing, as many have pointed out, they are just guidelines. With how many creatures there are bound to be some interesting examples of how creatures can work together, but I think that changing a creatures XP based on who it was fighting with might get unwieldy quickly.

However, I would argue that the difficulty of an encounter almost always scales up faster than linearly based on how many creatures you add regardless of which creatures are used. Enough kobolds for a moderate encounter for a higher level party will quickly get out of hand for the party. There will be so many that even a few fireballs will not take them all down, and with pack tactics, there will be plenty of 20s rolled.

As for the intellect devourer example with 10-20 creatures, it would be difficult for that group to surprise anyone. Either some of them roll low on their stealth check or someone has a ridiculous passive perception. Our rogue/monk has a passive perception of 19 at level 5. The party should realize that 15 creatures need more than a run in with guns blazing strategy. The intellect devourers (those that survive the wizard's alpha strike) need to get someone to fail a DC12 and roll average or higher on its 3d6 intelligence attack. At this point, the cleric probably has greater restoration, and the wizard can keep slinging fireballs or setting up walls of fire in choke points while the fighters use their multiple attacks to hit the low AC low HP enemies from afar.

Even for a level 20 party of 4, a group of 15 CR2 creatures (27000XP) is beyond a moderate (22800) encounter, and it only takes 1 intellect devourer to be successful to set up the next for a body swap, but at level 20 where this scenario starts to make senes, the wizard had better have Wish prepared, or the cleric has True Resurrection prepared. Any group of 15 CR2 creatures is something worth running away from or at least having a decent strategy before engaging. Even 15 ogres, with their higher to hit modifier, damage and hp could be a real threat.

While the guidelines are not perfect, they have gone through play testing and are useful as guidelines. In some of these more extreme encounters, a playtest using copies of your character's sheets before the game might be worth running. That way, you know if the encounter will be too deadly. From all of my experience so far, the guidelines give a good indication of the difficulty of an encounter and how many of the party will be face down on the floor.

 

[keterys]

but I think that changing a creatures XP based on who it was fighting with might get unwieldy quickly.

Except that's what we're already doing, by doing the "if there's more, multiply it"

In _most_ cases, my experience has been that the XP should scale linearly -or- it should actually be considered easier. At higher levels, it's really easy to kill hordes of creatures, use spells or environment to bottleneck, have things just have _so much trouble_ hitting the primary tanks, etc.

The intellect devourers (those that survive the wizard's alpha strike) need to get someone to fail a DC12 and roll average or higher on its 3d6 intelligence attack.

It's doesn't take rolling average or higher to roll higher than _0_ on 3d6.

Even 15 ogres, with their higher to hit modifier, damage and hp could be a real threat.

This really isn't my experience, but experiences can vary, I'll admit.

While the guidelines are not perfect, they have gone through play testing and are useful as guidelines.

Those guidelines did not go through a wide playtest, at all. None of the monsters went through a wide playtest, at all, when you get down to it - the public playtests all dealt with horrible far too easy monsters, and never had guidelines at all for building encounters.

So far, the best guideline for me has been total xp for the day, because that shows how many resources people have probably gone through. CR points out troublesome swingy fights, and application of resources largely solves any difficult fight, it's being caught without those resources that leads to trouble.

 

[Tormyr]

Except that's what we're already doing, by doing the "if there's more, multiply it"

I am sorry, I thought you were referring to different XP for a creature based on which kind of allies it is fighting with, and that is what my answer was referring to instead of the existing multiplying by number of creatures.

In _most_ cases, my experience has been that the XP should scale linearly -or- it should actually be considered easier. At higher levels, it's really easy to kill hordes of creatures, use spells or environment to bottleneck, have things just have _so much trouble_ hitting the primary tanks, etc.

It can be easy, but in the ogre example they could survive a couple fireballs or other defenses and charge through to the back lines.

It's doesn't take rolling average or higher to roll higher than _0_ on 3d6.

??? The Devour Intellect action requires the target to make a DC12 Int saving throw. If it fails that, the ID then rolls 3d6 to meet or beat the targets Int score, not its modifier. The average int for a character is probably around 10. The average roll of 3d6 is 10.5. The Body Thief ability requires an incapcitaed target to initiate an Int contest with. For someone who has been ability drained to 0 from Devour Intellect, they have to beat the ID's roll by 6 to survive.

This really isn't my experience, but experiences can vary, I'll admit.

Totally agree.

Those guidelines did not go through a wide playtest, at all. None of the monsters went through a wide playtest, at all, when you get down to it - the public playtests all dealt with horrible far too easy monsters, and never had guidelines at all for building encounters.

So far, the best guideline for me has been total xp for the day, because that shows how many resources people have probably gone through. CR points out troublesome swingy fights, and application of resources largely solves any difficult fight, it's being caught without those resources that leads to trouble.

I think they playtested these monsters internally, learned how difficult different abilities were and used those to create the creatures and their CR. Hopefully we get those guidelines in the DMG. Granted, I do not know this for a fact, but it is what I would have done.

 

[keterys]

It can be easy, but in the ogre example they could survive a couple fireballs or other defenses and charge through to the back lines.

Ogres don't really have ways of charging through back lines unless it's a wide open field (ie, the PCs have not controlled the fight), nor of coping with mass difficult terrain (such as from spell effects), or status effects.

The Devour Intellect action requires the target to make a DC12 Int saving throw.

Sorry, I'm confusing the issue by referring to Body Thief, not Devour Intellect. It's probably true that it's better for other creatures that stun, like mind flayers, to keep around (to turn stunned into dead), but hey, mass attempts can still be pretty effective.

I think they playtested these monsters internally, learned how difficult different abilities were and used those to create the creatures and their CR. Hopefully we get those guidelines in the DMG. Granted, I do not know this for a fact, but it is what I would have done.

Yep, that's why I said they didn't playtest any of these things widely. A group here or there, absolutely, but hardly enough that I'd have faith in anything over level 6 or so being robust. The XP guidelines in particular the multiply for extra creatures I'm not sure saw any playtest by anyone, in any way. They are a theorycraft construct used for theorycrafting what might be reasonable for your group. They can get it right sometimes, but there's no rigor involved.