Encounter balance and CR

[jack]

There are some fairly sensible rules of thumb for "how challenging is this monster". Things like "this amount of hit points and armour class are roughly equivalent to this amount of hit points and armour class" and "if it has good saves, treat the effective hp as this much higher". And the same for attacks, and, how to use monsters with attacks stronger than defences and vice versa, and how not to depart too far from equivalence or you get monsters that are really boring (if they odn't do much damage but take forever to kill) or really swingy (if they do lots of damage but are very fragile).

But it seems like Dungeon Master's Guide always makes a dog's breakfast of explaining these. It presents a bunch of rules as a rigid algorithm and says "you can tweak it", whereas I feel like someone who understands the rules of thumb could have provided a template beginners could use. I may try to write that up, but in practice GMs usually use a lot of intuition to tune monsters and I may not have enough experience with 5e yet.

The encounter difficulty is also confusing. The DMG has a system which works, but gets even me bogged down in layers of arithmetic I don't quite get even when I understand the purpose. One of the expansion products has a simpler system, basically a chart of "one PC at this level can handle this many goblins in a typical encounter, or this fraction of a higher level monster", so it's easy to eyeball "ok, I've got three players so six goblins is reasonable". But then, harder and easier challenges come down to "assume there are 50% more players or 50% less players".

But no-one seems to have a good guideline for how many encounters per day make a reasonable experience. Ideally that's enough that parties aren't rewarded for always blowing their biggest attacks in the only big combat of the night, but not enough to slow the game down. But it seems like most groups evolve into different status quos. And there's lacking decent guidelines around, some groups have three "deadly" encounters, some groups have less fights and don't worry about using up per day abilities as much, some groups have this many little fights.

Also, looking at how monsters are calculated. 5e still assumes that monsters fit into a framework of "hp = N x d6+con for small creatures, N x d10+con for big creatures, etc", and that all their attacks are derived from their abilities. But really... the CR is more important than any of that. Having a system to say "if a monster gets hit by a strength enhancement, what happens" is necessary for DnD to stay interestingly open-ended. But monsters never really level up, or if they do, it's not important that they do so in a consistent way, so you should just choose hp and ac that seem plausible to the creature and about the right amount of challenge, and not pretend the underlying hit dice system adds anything. That would also make designing monsters simpler.

I think 4e went too far this way, in providing guidelines-that-were-treated-like-rules about how to treat minion monsters, boss monsters, etc, and basically saying players and npcs use completely unrelated rules. But it provided a lot of good ideas for handling monsters, of which 5e keeps several, and GMs privately use others behind the scenes. Things like each type of monster having a signature mechanical ability that distinguishes them: that always existed for, say, displacer beasts, but now it applies to wolves (pack tactics) and goblins (some rogue stuff) too.

Comments

[tigerfort]

I would argue that "how many encounters per day make a reasonable experience" is something almost entirely dependant on the specific group. Some players will want their characters to save every casting/ability for the most crucial moment and still finish a day tottering on their last legs, but others will get more fun from flinging high-level spells at individual goblins and camping when the cleric still has half their heals left. (And everything in between, obviously.)

It might be useful to have two sets of guidelines, one for (groups close to) each end of the scale. But most GMs would still wind up needing to work out the rules of thumb to use for their particular players by trial and error.

[jack]

I think this is entirely true. I wasn't imagining a single number, but more like you describe in the last paragraph, a range of what's typical, with GMs knowing that their party is usually on the "low number of encounters and the PCs don't optimise their tactics overmuch" end of the scale or the "every encounter will be rated deadly else the PCs won't break a sweat" end of the scale, but can still use the standard evaluations to compare unfamiliar scenarios and see if they stack up the same way as their usual ones.

[seekingferret]

I also think that 5E leans further back toward 1E in trying to say it's okay to design unbalanced encounters. The flatter level progression means that with some luck, players can defeat encounters above their level, and also that encounters below their level can sometimes be challenging. But also, having encounters that players need to skip or deal with in non-combat ways because the enemies are too powerful makes dungeons feel more immersive and alive and not just a series of designed encounters.

I think this drives some of the abstraction of the CR system in 5E. In 4E if you don't design your encounters exactly right, combat quickly becomes unfun, but that's not so much the case in 5E I find, so getting the CR ratings perfect is not as important. They're a guideline, not a rule.

[jack]

Good point. I'm not completely sure, but I do really quite like that: you can still stick to middle-of-the-road encounters if you like, but you *can* use a mix and expect the players to muddle through the occasional too-hard encounter, or learn which encounters to avoid.

And yeah, 5e will probably work well enough without (I mean, I think 3.5e was just as much of a mess, and 4e made things more systematic only by rigidly enforcing a fairly specific level of opponents), but I feel like it would still really benefit from some easy guidelines for an *average* challenge, even if you usually want a fairly wide variance around that.

[silveradept]

Rigidity of numbers seems like something that's part of stereotypically terrible role playing games, and it's not that encouraging for a system to not be able to get out of its own numbers and go forward from there. It at least be able to abstract it down to something that a first-time GM can use.

[jack]

Yeah, definitely. I think it's not been a prominent problem in practice for DnD 3.5e or 5e, because you can mostly ignore the monster creation guidelines if you use existing monsters, tweak existing monsters, or have the experience to just eyeball your own, and if you play with the same group regularly you'll usually know what works for them better than the standard guidelines would work anyway. Reading around, it seems like they did quite a lot of testing of the published monsters to get them right, but monster creation guidelines weren't really tested for some reason, and that does trip people up if they try to use them :(

[rysmiel]

But monsters never really level up, or if they do, it's not important that they do so in a consistent way,

I far prefer the paradigm in Pathfinder where monsters do indeed level up in a consistent way, even though it's more maths work than many people want to do, and am a good bit less than happy that this looks like being lost in Pathfinder 2.0.

I believe I have talked before about very much favouring giving players a wide range of encounters from much harder to much easier that they can handle in combat, as either of those are ways to convey information at a higher level of abstraction. And if one must do XP rather than levelling up at appropriate points in the campaign arc, weighting story and role-playing rewards much more heavily and combat results less so seems to be good for the style of game I favour.