D&D 5E Question about philosophy behind CR

[intently]

Does anyone know why the decision was made to design CR the way it is? Why is a CR X monster intended to be suitable for a party of level X adventurers? If I had been designing CR, I would have built it so that one CR X monster was a match for a single level X character, and a party of level X adventurers would be matched against a party of CR X monsters.

 

[Fanaelialae]

D&D 4th edition designed monsters that way, but the baseline assumption of that edition was that (barring solo and minion types) there would be roughly one monster for each PC in a given encounter. Which was fine, but it relied on using solo and minion types to create larger or smaller encounters, and those creature types didn't sit well with some people for a variety of reasons.

You can basically think of CR as a safety measure in 5e. It roughly indicates that a creature of CR X shouldn't be overpowered for characters of level X. A creature whose CR is greater than X might still be defeated by an X level party, but there's a greater chance that it has abilities that the party could have a hard time coping with (such as a monster that is likely to take a full hp fighter to 0 hp in a single round if it hits with all attacks).

 

[Tony Vargas]

Does anyone know why the decision was made to design CR the way it is? Why is a CR X monster intended to be suitable for a party of level X adventurers?

It's only approximately that according to the latest guidance, IIRC. It is also, probably not coincidentally, how 3e, which introduced CR, did it as a default. (Though it did more than one monster in a somewhat simpler way - two monsters of CR-2 were 'equal' to a single CR monster. So if you wanted a party of 4 to fight an equal number of monsters, the baseline would be 4 monsters of CR -4. That is simpler than 5e's encounter budgets and multipliers, IMHO, though now that I type it, maybe not vastly so, especially if working out encounters for a party of 5, or wanting to outnumber a party by some less convenient ratio.)

If I had been designing CR, I would have built it so that one CR X monster was a match for a single level X character,

Actually, in 3e that is how it worked. If that hypothetical 5th party member turned on his 4 allies, that'd be a 'standard' encounter - standard encounters were meant to be speed-bumps that incremented party resources down by as much as 25%, not even-money fair fights. (I mean, how it was supposed to work, obviously the CoDzilla turning on you would be different from the fighter doing so.)

and a party of level X adventurers would be matched against a party of CR X monsters.

Well, that's how 4e did it, and 5e was in part motivated by a violent anti-4e reaction, so it couldn't really take anything straight from 4e if it could possibly be avoided. Another side of that coin is that 5e had a mandate for 'fast combat' without resorting to using minis or any sort of play surface, so reverting to a tactically simple or 'static' 3e style combat (defaulting to PCs dog-piling a lone monster) also made sense, from there the 3e default CR structure also had to have made a certain amount of sense.

It didn't exactly dovetail neatly with Bounded Accuracy, though, so a comparatively simple formula for multiple monsters couldn't be devised, and we got what we got. :shrug: It's also not terribly dependable, so, rather than lose sleep over it, I say just 'design' combats by feel, you will, if you don't already, develop a feel for what it takes to challenge a party of D&D characters - or even just go full-on "status-quo"/sand-box style and don't design encounter to be a certain level of challenge, at all, just design the setting and let the chips fall where they may, it's also a legit DMing style, and players will adapt to it after a sufficiency of TPKs.

Which was fine, but it relied on using solo and minion types to create larger or smaller encounters.

The simple numeric way that 4e did minions & solos (and elites) wouldn't have played with BA, but, ultimately, much-lower-CR monsters serve the same function as minions (if even less durable, since they can die to a failed save:1/2), it's just more complicated a substitution than a simple 4:1 ratio. And, 5e Legendary monsters are essentially designed to be Solo-worthy, even delicately lifting some tricks from later 4e and filing the serial numbers off to get there.

With BA, though, a normal monster can serve as a 'standard' or 'minion' or even a sort of elite or poor 'solo' depending on it's CR relative to the party, and a Legendary one can, similarly, play like an elite if lower level than the party, or Solo if of similar level. That was part of the point of BA, to let one monster stat block serve the DM across a number of levels.

 

[Fanaelialae]

Actually, in 3e that is how it worked. If that hypothetical 5th party member turned on his 4 allies, that'd be a 'standard' encounter - standard encounters were meant to be speed-bumps that incremented party resources down by as much as 25%, not even-money fair fights. (I mean, how it was supposed to work, obviously the CoDzilla turning on you would be different from the fighter doing so.)

I didn't find 3e CR to be a terribly useful metric. 99% of the time I'm not interested in what has a 50/50 chance of killing a PC if he fights it one on one, nor what will use 25% of the party's resources if they fight it 4:1.

I liked 4e's approach, where a level X monster was a challenge for a level X PC but wasn't likely to beat the PC unless perhaps the PC was low on resources. That's what I think of as a "match" for a character, because it's typically what I want to use.

5e's system is a bit more involved (which I why I created a spreadsheet to make it easy) but it generates reasonable results (as long as you keep in mind the actual DMG definitions of easy/med/hard/deadly).

The simple numeric way that 4e did minions & solos (and elites) wouldn't have played with BA...

Sorry, but this doesn't make much sense to me. I can pretty easily modify existing 5e monsters into solos/minions by increasing/decreasing their hp and modifying the CR accordingly. Bounded accuracy has nothing to do with it as far as I can see. If the CR changes by enough to increase/decrease the creature's proficiency then its attack and DCs will be off slightly from what they ought to be, but that's only an issue if you need the math to add up (the system itself doesn't care).

 

[dropbear8mybaby]

The "philosophy" was basically, "Let's eyeball all these and then make up some sort of convoluted system at the last minute before publishing and claim that we were using it all along, haha."

 

[Tony Vargas]

I liked 4e's approach, where a level X monster was a challenge for a level X PC .. That's what I think of as a "match" .

Sure, it was intuitive enough that way.

(as long as you keep in mind the actual DMG definitions of easy/med/hard/deadly).

pointless/trivial/easy/swingy? ;P

Sorry, but this doesn't make much sense to me. Bounded accuracy has nothing to do with it as far as I can see.

Yeah, I was not clear with that one. BA is a simple concept but multifaceted, one thing it accomplished is letting the DM use the same monster stat block for the same monster at different levels. Monsters designed as 'Minions' for their face value CR would have undercut that, so the more complex guidelines are used, instead.

 

[FormerlyHemlock]

It's only approximately that according to the latest guidance, IIRC. It is also, probably not coincidentally, how 3e, which introduced CR, did it as a default. (Though it did more than one monster in a somewhat simpler way - two monsters of CR-2 were 'equal' to a single CR monster. So if you wanted a party of 4 to fight an equal number of monsters, the baseline would be 4 monsters of CR -4. That is simpler than 5e's encounter budgets and multipliers, IMHO, though now that I type it, maybe not vastly so, especially if working out encounters for a party of 5, or wanting to outnumber a party by some less convenient ratio.) Actually, in 3e that is how it worked. If that hypothetical 5th party member turned on his 4 allies, that'd be a 'standard' encounter - standard encounters were meant to be speed-bumps that incremented party resources down by as much as 25%, not even-money fair fights. (I mean, how it was supposed to work, obviously the CoDzilla turning on you would be different from the fighter doing so.)

Coincidentally, this is still basically true. A reasonably-optimized 5E character of level N usually clocks in just shy of CR N, and a solo PC of level N seems to do pretty well against creatures of CR N. E.g. Barbarian 2 vs. CR 2 Ogre is a pretty fair fight even if the Barbarian doesn't fight particularly smart.

5E tries by default to steer you toward Medium encounters ("speedbumps" as Tony puts it), which means it's basically okay with your party fighting one or two clones of the party (Medium-Hard difficulty) which have 1/16 to 1/4 of the party's full power, but a full clone of the entire party would be somewhere around triple-Deadly, and the game by default steers you away from that. But there's no reason you can't do it anyway. It depends whether your preferred DMing style is to provide a challenge that can be beaten through sheer brawn and the power of superior firepower, or a nemesis equal to or greater than the PCs in raw might who has to be beaten by player skill and superior tactics. (I suppose there also exists a DMing style which is all about challenging the PCs with foes who have vastly inferior firepower but superior tactics--the Tuckers' Kobolds style. Be careful with that one though because it risks becoming adversarial DMing.)

 

[Fanaelialae]

Yeah, I was not clear with that one. BA is a simple concept but multifaceted, one thing it accomplished is letting the DM use the same monster stat block for the same monster at different levels. Monsters designed as 'Minions' for their face value CR would have undercut that, so the more complex guidelines are used, instead.

I see. I don't consider the wide level range where a CR is usable to be an aspect of BA, so much as a result of this particular implementation therefore (it's not a cause, it's an effect). 4e had a very different kind of BA (relative BA, which ensured that values would fall within a certain range for certain levels) which rendered monsters only usable for a narrow range of levels, but was nonetheless fundamentally a BA system IMO.

 

[OB1]

Which makes sense, since fighting a clone of the party would result in a 50/50 outcome decided solely by the dice assuming the same level of player skill on both sides.

A deadly encounter in 5e is one that can be deadly for the party due to poor decision making or poor dice rolls, but that should always be beaten with average rolls and tactics equal to the DM.

 

[Sword of Spirit]

It's also not terribly dependable, so, rather than lose sleep over it, I say just 'design' combats by feel, you will, if you don't already, develop a feel for what it takes to challenge a party of D&D characters - or even just go full-on "status-quo"/sand-box style and don't design encounter to be a certain level of challenge, at all, just design the setting and let the chips fall where they may, it's also a legit DMing style, and players will adapt to it after a sufficiency of TPKs.

Running by feel is usually a strength of mine, but it has taken me forever to get it for 5e combat challenge (as opposed to setting ability check DCs on the fly, which I love doing), and I'm still not 100% on it.

One thing that I tried in examining some past encounters (my friend has taken over DMing for a few weeks, so haven't had a chance to use it yet) is to do this:

1) Determine what the party's average damage output is in a round, assuming all attacks hit.
2) Determine the party's total "health resources." Basically, sum the hit points, and then add things like Second Wind and Dark One's Blessing.
3) Do the same for the opponents.
4) Assume the party hits more often than they miss (because they usually do), and the monsters miss more often than they hit (this might change at higher levels, but it's accurate at low level), and then see how the numbers compare.

It worked out remarkably well at predicting the outcome of the recent fights. They ran about as many rounds as it seemed like they should by the math. Basically, I wouldn't have bothered with steps 1-3 in prior editions, but 5e's numbers are sometimes unexpected (the party's numbers were significantly higher than I thought they were). So when I took those extra steps, I was then able to apply my normally good gut feelings in step 4 and make proper estimates.