Two Sets of Rules, the Good and Bad?

[MaxWilson]

He, and it is clearly indicated by the little blue circle/arrow symbol under his name, just like the one under my name.

Oh, interesting. Thanks.

[KorvinStarmast]

Oh, interesting. Thanks.

That symbol (and the circle/cross symbol for lady) has been around since the late 60's as far as I can recall, maybe earlier. (Yeah, I am that old...) I vaguely recall a loose association with the feminist movement of those days, but that's a very rough recall so don't quote it as gospel.

[Sorinth]

What do you mean by "add features"? The feature is there in the PHB. It probably won't be relevant, and I'm certainly not going to write "Divination Savant" in my notes because "Diviner 7" already tells me that it has all of the features of a Diviner.

Eliminating Divination Savant or Expert Divination would be extra work, because then I'd have to write "Diviner 7, but without Divination Savant", and why would I ever do that? It complicates the NPC for no reason. I need that space for other things like documenting the NPC's relationships with PCs and other NPCs.

If just writing the subclass and level is enough for you to remember every feature at the NPCs disposable without having to waste time looking it up then I can see why that would work for you. I know for myself it's a sure fire way to forget to use abilities or be forced to look them up to get the exact wording/details. So I'd want Portent on the sheet so that I don't have to try and remember whether there there was a range limit like 60' or not but I wouldn't put Divination Savant because it's not an ability that is worth the space because like you said the space can be better spent.

[Mikal]

That symbol (and the circle/cross symbol for lady) has been around since the late 60's as far as I can recall, maybe earlier. (Yeah, I am that old...) I vaguely recall a loose association with the feminist movement of those days, but that's a very rough recall so don't quote it as gospel.

Though note the symbol doesn’t appear if you’re on the mobile site

[MaxWilson]

That symbol (and the circle/cross symbol for lady) has been around since the late 60's as far as I can recall, maybe earlier. (Yeah, I am that old...) I vaguely recall a loose association with the feminist movement of those days, but that's a very rough recall so don't quote it as gospel.

It's much older than that (Wikipedia says the convention was introduced by Linnaeus in the 1750's) but I never realized before that it showed up on the site. My ad blocker is set to block everyone's avatars, and I never noticed before that there's another little gender icon that it's not blocking.

If just writing the subclass and level is enough for you to remember every feature at the NPCs disposable without having to waste time looking it up then I can see why that would work for you. I know for myself it's a sure fire way to forget to use abilities or be forced to look them up to get the exact wording/details. So I'd want Portent on the sheet so that I don't have to try and remember whether there there was a range limit like 60' or not but I wouldn't put Divination Savant because it's not an ability that is worth the space because like you said the space can be better spent.

Yeah, for the NPCs I create it is***.

*** I couldn't tell you off the top of my head what a Beastmaster's 17th level ability is, but I don't create 17th level Beastmaster NPCs. If I did I'd just look it up as part of the design process, and then I would write down a note if it was relevant and I thought I might forget it. P.S. Oh! Apparently the 17th level ability is... nothing (except for the obvious: another level of spells).

Anyway, sounds like we basically agree about how to do NPCs, although I wouldn't bother writing down Portent either unless I was going to pre-roll the Portent dice and write that down, e.g. "Portent 17, 5". But I probably wouldn't even do that until something came up that made Portent potentially relevant, like the NPC casting a Contact Other Plane spell or a fight starting.

[Tanarii]

42 HP: base defensive CR 1/4.

For a base offensive CR of 9. Level 7 ==> +4 INT ==> save DC = 15, which is one below the specified, so final offensive CR is unchanged = 9.

Total CR: 4.5 ==> 5 (rounding up due to damage output). So I was off a bit. But functionally, it acts like a really bity monster of much lower CR, because it will die if a PC sneezes on it.

You don't treat fractional CRs as a decimal point to be averaged. CR 1/4 is 2 steps down from CR 1. The average of CR 9 and CR 1/4 is CR 4. Five steps up from CR 1/4 and five down from CR 9.

Does it really take you 5-10 minutes to write down "Diviner 7"? Or is it the spell list that takes you 5-10 minutes to write down? You need that spell list no matter which approach you use to NPC creation.

I think now we know why NPCs with classes are so hard for you, PhoenixPhyre: you're doing unnecessary work.

And it's just as easy to write down Diviner (Volvo pg ##).

The only difference is you appear to have the very complex PC build memorized (as do I for many of them), but not a Goblin or Orc, or more understandably a not-common NPC stat block. A stat block takes roughly as much time to jot down and invent as does custom notes of the important stuff you plan to need from a PC build, and customizing either for the particular NPC takes extra time regardless.

-----

Personally I'd rather not use PC stat blocks for NPCs exactly because they are designed for resource attrition across an adventuring day. Even in CaW scenarios, they usually aren't anywhere near the appropriate use-case for the rule set when used for enemies.

They are the wrong tool for the job.

[Telok]

An analogy: Someone making a Skyrim mod could build all his own animations and scripting from scratch. Or, realizing that he's just making a minor change to an existing thing, alter the included asset and scripts and call it a day. There are cases where the first is the right thing to do. But they're cases that the vast majority of modders will never run into.

Ah, a surprisingly apt analogy. See, both the Skyrim mob and your critter out of the D&D MM are just a text and numbers that get interpreted through a rules engine (in this case the DM equates to the graphics card).

So take, say, a rat. What's it got? The usual ac, hp, dmg, etc. numbers. No problems there, same stuff acrossthe crpg & d&d. Oh, a scurry animation/default tactics, a rat model/description, and a disease effect on attack/save vs disease . Ok, it's a rat. You want a simple change? How about a better rat that's poisonous? Add some numbers, increase the size 20%, swap disease effect for poison effect.

You're right, no need for complicated stuff there in any system. How about a two headed web shooting snake instead? Oops, we'll need some new models, new animations, scripts for two heads and slithering instead of scurrying. Oh well, how about in d&d? Well two heads might do something to mental saves, there's no good cr estimate for the webs because they aren't hp damage, and it probably shouldn't fight like a rat. Well, looks like were pretty much writing stuff up from scratch no matter what system we use.

CRPGs evolved from gamers who learned to code and wanted RPGs. They are an RPG rule engine just like TTRPGs. They just swap an inflexible calculator and graphics card for the human DM.

Building monsters differently than PCs is fine. Nobody expects parity between a halfling thief and an ancient red dragon. Non-parity is sort of the point there. Building NPCs is less perfect. If you just need a random thug or an insane sorcerer that sold brain & soul for forbidden magic? Yeah, you can skimp all the rules you like. You just need the combat stats for a fight. But if the party takes an enemy NPC knight captive and the PC fighter wants to turn then into an ally and learn that special knight move... Well it gets awkward if they're too different. 4e had... notable issues... around that sort of thing. It gets pretty lumpy with enemies on cr appropriate riding animals too.

[MaxWilson]

Personally I'd rather not use PC stat blocks for NPCs exactly because they are designed for resource attrition across an adventuring day. Even in CaW scenarios, they usually aren't anywhere near the appropriate use-case for the rule set when used for enemies.

Are all NPCs enemies? Are all enemies designed for a single short scene?

Not in my games.

[Tanarii]

Are all NPCs enemies? Are all enemies designed for a single short scene?

Not in my games.

No, but I use PC rules for henchmen. So that's a fair point. And I would for allies too, although that wasn't relevant to my campaign.
(A good example of your point, the allies in OotA are monster stat blocks, and pretty useless because of it.)

And let's put it this way, all enemies are FAR closer to single short scene use case scenario than they are to an adventuring day (or more) use case scenario. Playing a CaW game or CaS game won't change that.

And if they're neither, I don't need a stat block. And if an enemy becomes an ally or vice versa, I'd much rather convert their stat block to the next closest thing that use the other kind. (Although that's a lot of extra work.)

[MaxWilson]

No, but I use PC rules for henchmen. So that's a fair point. And I would for allies too, although that wasn't relevant to my campaign.
(A good example of your point, the allies in OotA are monster stat blocks, and pretty useless because of it.)

And let's put it this way, all enemies are FAR closer to single short scene use case scenario than they are to an adventuring day (or more) use case scenario. Playing a CaW game or CaS game won't change that.

That's interesting. For me, many but not all enemies are a presence for a short time, but others last much longer. E.g. the golems guarding an arcane vault that the PCs might like to break into may last for in-game YEARS. A hobgoblin army may last months. A rival NPC (like Belloq to Indiana Jones) may also last in-game years, especially if the relationship is such that he can't easily be disposed of. (Either he represents The Law, or murder is simply illegal, or he's a scion of a house you need to stay in good graces with, or he's your brother-in-law... C.f. Knife Theory, https://www.gmbinder.com/share/-L-9CvlTWhoADagJfSZO)

Generally, if someone weren't intended to have a presence that lasts for at least several scenes (even if they are offscreen), I wouldn't bother making them a human NPC, I'd just use a monster. Exception: minions attached to an important NPC might or might not make sense as low-level Fighters (no subclass) instead of monsters (orcs, ogres, trolls, etc.), but if I really want the players to know it's okay to ruthlessly kill the minions once open conflict occurs I'll make them monsters.

An off-screen NPC still shows signs of activity. Witnesses interviewed, Fireballs launched, allies recruited, treasure stolen, ribald graffiti painted, etc. Even if they eventually die in one Action Surged lucky attack, they still will have had an impact on the play experience, or they wouldn't be worth making as NPCs in the first place.

[PhoenixPhyre]

Ah, a surprisingly apt analogy. See, both the Skyrim mob and your critter out of the D&D MM are just a text and numbers that get interpreted through a rules engine (in this case the DM equates to the graphics card).

So take, say, a rat. What's it got? The usual ac, hp, dmg, etc. numbers. No problems there, same stuff acrossthe crpg & d&d. Oh, a scurry animation/default tactics, a rat model/description, and a disease effect on attack/save vs disease . Ok, it's a rat. You want a simple change? How about a better rat that's poisonous? Add some numbers, increase the size 20%, swap disease effect for poison effect.

You're right, no need for complicated stuff there in any system. How about a two headed web shooting snake instead? Oops, we'll need some new models, new animations, scripts for two heads and slithering instead of scurrying. Oh well, how about in d&d? Well two heads might do something to mental saves, there's no good cr estimate for the webs because they aren't hp damage, and it probably shouldn't fight like a rat. Well, looks like were pretty much writing stuff up from scratch no matter what system we use.

CRPGs evolved from gamers who learned to code and wanted RPGs. They are an RPG rule engine just like TTRPGs. They just swap an inflexible calculator and graphics card for the human DM.

Building monsters differently than PCs is fine. Nobody expects parity between a halfling thief and an ancient red dragon. Non-parity is sort of the point there. Building NPCs is less perfect. If you just need a random thug or an insane sorcerer that sold brain & soul for forbidden magic? Yeah, you can skimp all the rules you like. You just need the combat stats for a fight. But if the party takes an enemy NPC knight captive and the PC fighter wants to turn then into an ally and learn that special knight move... Well it gets awkward if they're too different. 4e had... notable issues... around that sort of thing. It gets pretty lumpy with enemies on cr appropriate riding animals too.

Why do I need to calculate things into CR which don't factor in? Why do I care? And there's already good analogs for all those things. Grab the webs from a spider, slithering is narrative, and two-headed is already a trait. Done. Total time: 30 seconds, maybe a few minutes if we want to calculate all the pieces.

And for NPCs...uh, wat? You use the stat blocks and add in things as needed. Their non-combat stuff is just entirely what they should have for their role in the world. No need to worry about "well, he's a wizard, so he can't have XYZ as a skill unless he's taken those three PRCs in that particular order, and then..." It's way better at having world-fidelity than a more crunchy system where you have to explain how you got to all the values. Because you're not trading off things that have to be balanced for PC use (which inevitably come with game oriented tradeoffs and requirements that don't always make sense for that particular person.

And as for dissonance...I've never felt it to be all that difficult to give people boons as quest rewards. You want to learn from him? Do a favor and sure, the ability is yours. Easier, cheaper, and less unbalancing than giving magic items as rewards IMX.

-----------

I think the big disconnect is that I don't consider CR to be more than it is. A first-pass tool, mainly for new DMs who don't know how to judge difficulty for their particular party yet. And it does OK at that. I don't bother calculating CR or the encounter budget any more, because I have a sense of what the party can take and I'm just building based on what's in the world. CR is not, nor was it intended to be, the be-all and end-all of difficulty. It's not supposed to take into account highly situational factors (such as web or two heads, both of which can be devastating or pointless depending on the group, the other enemies, the terrain, etc). So stop asking it to take those into account. Use it for what it's designed for, under the assumptions it was designed for.

I made a group of "former adventurers turned undead" for a session tonight. One was stock, just pulled a deathlock wight. The other two were wights, but one with plate armor and a shield + action surge and the other with sneak attack and a magical dagger/magical armor. Oh, and bumped the HP on them a bit. Total build time? about 3 minutes for the entire encounter. And most of that was programming it into the VTT I have to use because the sessions are remote. If it'd been in-person, I'd likely have not even written stat blocks and done the translation in my head.

I've got another one coming up with a fancy construct. Grabbed a Gearforged Templar (Tome of Foes IIRC), said "great, his glaive does force damage[1] and he's got a short-range teleport 1x/day". Done. Total time? 10 seconds. No rebalancing needed. Frees me up to plan the stuff that actually matters (which is very very very very rarely the mechanical bits and bobs). Especially since the other way involves more work (have to come up with all the numbers from scratch, have to dig through books to see what's available to whom when, etc).

If I wanted a caster who could cast those 3 spells off of that list and these other 4 spells off of the other list, all with CHA (for various in-world reasons), done. No extra work. I write them down and be done with it. Instead of having to figure out what multi-class would be required, and not be able to do it.

[1] They's magical constructs.

[Tanarii]

That's interesting. For me, many but not all enemies are a presence for a short time, but others last much longer. E.g. the golems guarding an arcane vault that the PCs might like to break into may last for in-game YEARS. A hobgoblin army may last months. A rival NPC (like Belloq to Indiana Jones) may also last in-game years, especially if the relationship is such that he can't easily be disposed of. (Either he represents The Law, or murder is simply illegal, or he's a scion of a house you need to stay in good graces with, or he's your brother-in-law... C.f. Knife Theory, https://www.gmbinder.com/share/-L-9CvlTWhoADagJfSZO)

I don't see how that impacts that the NPC is using something close to an adventuring days worth of resources in the process of opposing the PCs and anything else they're doing in the day, they're far better built on the monster stat block model.

One thing that I've done before is assume some kind of resources expended for a specific encounter, and if that impacts CR recalculate it. That applies to monsters too, especially ones with spell lists that have used their highest damaging spells.

[MaxWilson]

And for NPCs...uh, wat? You use the stat blocks and add in things as needed. Their non-combat stuff is just entirely what they should have for their role in the world. No need to worry about "well, he's a wizard, so he can't have XYZ as a skill unless he's taken those three PRCs in that particular order, and then..."

Those aren't even actual rules for any wizards, including PCs. Sounds like you're just carrying over opinions from 3E to 5E without updating them based on the actual rules for 5E.

I don't see how that impacts that the NPC is using something close to an adventuring days worth of resources in the process of opposing the PCs and anything else they're doing in the day, they're far better built on the monster stat block model.

One thing that I've done before is assume some kind of resources expended for a specific encounter, and if that impacts CR recalculate it. That applies to monsters too, especially ones with spell lists that have used their highest damaging spells.

Spoiler: My original response

Show

You snipped the part that talks about how offscreen NPCs still use resources in pursuing their goals, but to reiterate and elaborate: scorch marks from a Fireball and corpses mark where the NPC expended spell slots on Fireball; allies such as grungs may result from the NPC casting Tongues, Suggestion, or Enhance Ability; bloodstains may mark the spot where an NPC took an opportunity attack stealing a precious artifact; an NPC who's actively opposing you may send a steady stream of demons your way while hiding in a "safe" location, but if you catch him in that safe location he'll be down those spell slots unless he's been resting long enough to regain them.

(Honestly an NPC who specifically opposes you can put a ton of hurt on you just by encouraging monsters to not invite defeat in detail--a 3rd level chainlock with a Sprite familiar who remotely persuades two group of five orcs to merge into one group of ten orcs has doubled their combat power against the PCs.)

Bottom line: significant NPCs should have an agenda, even while they're offscreen. If their presence in the story is limited to a few seconds of combat only, they might as well just be monsters (like those ten orcs) instead of NPCs.

Edit: Now that I reread what you wrote I'm concerned by the fact that I don't understand what you mean by "anything else they're doing in the day, they're far better built on the monster stat block model." How and in what way? What point are you trying to make here?

[OldTrees1]

I think the big disconnect is that I don't consider CR to be more than it is. A first-pass tool, mainly for new DMs who don't know how to judge difficulty for their particular party yet. And it does OK at that. I don't bother calculating CR or the encounter budget any more, because I have a sense of what the party can take and I'm just building based on what's in the world. CR is not, nor was it intended to be, the be-all and end-all of difficulty. It's not supposed to take into account highly situational factors (such as web or two heads, both of which can be devastating or pointless depending on the group, the other enemies, the terrain, etc). So stop asking it to take those into account. Use it for what it's designed for, under the assumptions it was designed for.

That is a big disconnect.

At the beginning of 3E, I thought of CR as WotC's estimation of the average expected difficulty of that monster.
At the beginning of 3.5, I thought of CR (or more appropriately Encounter Level) as a useful term for the difficulty I expect the party to face from that monster or encounter. By then I still used WotC's estimation as a first pass baseline but I had experience adjusting the EL with ad hoc modifiers for different types of hazards, ambush, etc.
Deeper into 3.5, I thought of CR (or more appropriately Encounter Level) as a useful term to describe the difficulty this particular party would face, and I had the experience necessary to estimate that concept unless I went too crazy with custom abilities.

Basically I adopted the term for its professed meaning in 3E (rating the challenge) while acknowledging estimating difficultly was a task I needed to do.

However since I acknowledge estimating difficulty was a task I needed to do, I wanted WotC's first pass to be useful to newer DMs. I wanted the CRs to be WotC's first pass at estimating the challenge of the monster. If WotC can only check damage rather than attempt to address offense, that is a bit disappointing. They do a better job on defense where they account for saves in addition to effective hp.

I would be more disappointed if I had higher faith in WotC's estimation. I would be less disappointed if 5E had used a smooth power curve rather than the "tier boundaries must be special" which hampers my own estimation. Bounded accuracy also hampers my estimation, but if the "first pass tool for new DMs" handled that in a way that made sense when calculating encounters of mixed CR, I could have rebuilt my estimation abilities easier, earlier, and further.

[MaxWilson]

I would be more disappointed if I had higher faith in WotC's estimation. I would be less disappointed if 5E had used a smooth power curve rather than the "tier boundaries must be special" which hampers my own estimation.

For PCs, yes, there's a spike, especially at level 5. But there's no corresponding spike in power at CR 5 relative to CR 4, although of course there's a pretty big spike at CR 1 relative to CR 1/2. Monster power growth isn't aligned with tier boundaries.

[OldTrees1]

For PCs, yes, there's a spike, especially at level 5.

Exactly. Those abnormal spikes make it harder for me to retune my difficulty intuitions.

[PhoenixPhyre]

For PCs, yes, there's a spike, especially at level 5. But there's no corresponding spike in power at CR 5 relative to CR 4, although of course there's a pretty big spike at CR 1 relative to CR 1/2. Monster power growth isn't aligned with tier boundaries.

That's because, unlike 3e, there's absolutely no assumption that CR ~ level. For a "stock" party, the "safe"[1] range of CRs go from ~level + 3 down a level-dependent amount. Although there's a general presumption that you'll more frequently fight things below your CR than above, and that most fights will involve multiple opponents (my personal sweet spot is about 6 enemies to 5 players up to twice as many enemies as players).

So that CR 5 creature might be any of the following:
* A boss monster with some minions for a mid-T1 party.
* An "equal match" in groups of roughly the same size as a party of level 6-8s
* A minion for a low T3 group.

Some work with XGtE's encounter building guidelines suggests that the median non-boss enemy faced by a party of level 20s is roughly CR 10-11, in a range from CR 7 - CR 14 or so. In fact, CR ~ Level is the least likely to occur--too strong to face in significant groups, too weak for a boss.

[1] ie unlikely to simply evaporate while also unlikely to TPK the party on a bad roll by accident. The sort of thing you could theoretically face multiple times a day.

[MaxWilson]

Exactly. Those abnormal spikes make it harder for me to retune my difficulty intuitions.

Huh. So even though the monster power curve is somewhat smooth (two CR 6s are roughly interchangeable with four CR 3s), it still trips you up because you're not sure how strong the PCs are? That's a different problem than the one I thought you were describing.

That's because, unlike 3e, there's absolutely no assumption that CR ~ level. For a "stock" party, the "safe"[1] range of CRs go from ~level + 3 down a level-dependent amount. Although there's a general presumption that you'll more frequently fight things below your CR than above, and that most fights will involve multiple opponents (my personal sweet spot is about 6 enemies to 5 players up to twice as many enemies as players).

So that CR 5 creature might be any of the following:
* A boss monster with some minions for a mid-T1 party.
* An "equal match" in groups of roughly the same size as a party of level 6-8s
* A minion for a low T3 group.

Some work with XGtE's encounter building guidelines suggests that the median non-boss enemy faced by a party of level 20s is roughly CR 10-11, in a range from CR 7 - CR 14 or so. In fact, CR ~ Level is the least likely to occur--too strong to face in significant groups, too weak for a boss.

[1] ie unlikely to simply evaporate while also unlikely to TPK the party on a bad roll by accident. The sort of thing you could theoretically face multiple times a day.

While I disagree with many of the conclusions you reach here, and it's also not relevant to what I was discussing with OldTrees1 (how smooth scaling impacts estimation), I do realize that you thought you were being helpful by sharing your opinions and I thank you for your good intentions.

[OldTrees1]

Huh. So even though the monster power curve is somewhat smooth (two CR 6s are roughly interchangeable with four CR 3s), it still trips you up because you're not sure how strong the PCs are? That's a different problem than the one I thought you were describing.

Yeah, 3E really spoiled me by letting the PCs power curve and the monster power curve be smooth exponentials of roughly the same base. That means it takes seconds for me to design balanced encounters of mixed enemies and unequal ratios. Boss fight? Ambush? Guardians with traps? Swarm with coordinators? It was really easy which let me focus more time on making sure they were interesting compositions rather than fighting the system trying to calculate what mixed compositions wouldn't be a TPK.

My 3E earlier example of 1 CR 10, 2 CR 8s, 4 CR 4s as an EL 13 boss fight took ~30-45 seconds to calculate and double check.

Other fast examples:
EL 13: 3 CR 9 Guardians with a CR 7 hazard and 2 CR 5 traps
EL 13: 2 sets of (4 CR 6 Swarmlings with a CR 8 Commander)
EL 13: 2 sets of (8 CR 4 Swarmlings with a CR 6 Squad Leader) with 1 CR 8 Commander.
EL 13: An ambush with 2 CR 7 Hosers 2 CR 7 Artillery and 4 CR 5 Assassins.
PS: These are all TPKs in 5E. Don't try them.

Edit: If only we could combine that 3E ease with simple stat blocks like 5E.

[MaxWilson]

Yeah, 3E really spoiled me by letting the PCs power curve and the monster power curve be smooth exponentials of roughly the same base. That means it takes seconds for me to design balanced encounters of mixed enemies and unequal ratios. Boss fight? Ambush? Guardians with traps? Swarm with coordinators? It was really easy which let me focus more time on making sure they were interesting compositions rather than fighting the system trying to calculate what mixed compositions wouldn't be a TPK.

My 3E earlier example of 1 CR 10, 2 CR 8s, 4 CR 4s as an EL 13 boss fight took ~30-45 seconds to calculate and double check.

Other fast examples:
EL 13: 3 CR 9 Guardians with a CR 7 hazard and 2 CR 5 traps
EL 13: 2 sets of (4 CR 6 Swarmlings with a CR 8 Commander)
EL 13: 2 sets of (8 CR 4 Swarmlings with a CR 6 Squad Leader) with 1 CR 8 Commander.
EL 13: An ambush with 2 CR 7 Hosers 2 CR 7 Artillery and 4 CR 5 Assassins.
PS: These are all TPKs in 5E. Don't try them.

Edit: If only we could combine that 3E ease with simple stat blocks like 5E.

But 5E is so easy by default (if you avoid trick monsters and advanced tactics) that I can still do the same thing anyway in 30-45 seconds. Level 13 fight: how about one CR 13 Devourer, a couple of CR 5 Star Spawn Manglers, and let's add three CR 4 Flameskulls as magical support. Same 45 seconds to compute, similar level of certainty (it won't be a TPK unless the party is exceptionally weak but will be strong enough to feel dangerous/punish mistakes). Same flaws too: have to know the monster abilities, not just the CR, and lots of variance in the actual power level of said level 13 party based on builds and tactics.

CR just isn't very precise (the difference in stats between CR 6 and 8 is small, just look at white vs. green young dragons), so any variance from guesstimating difficulty is swamped by the variance in party builds/tactics anyway as well as dice variance.

[JNAProductions]

Yeah, 3E really spoiled me by letting the PCs power curve and the monster power curve be smooth exponentials of roughly the same base. That means it takes seconds for me to design balanced encounters of mixed enemies and unequal ratios. Boss fight? Ambush? Guardians with traps? Swarm with coordinators? It was really easy which let me focus more time on making sure they were interesting compositions rather than fighting the system trying to calculate what mixed compositions wouldn't be a TPK.

My 3E earlier example of 1 CR 10, 2 CR 8s, 4 CR 4s as an EL 13 boss fight took ~30-45 seconds to calculate and double check.

Other fast examples:
EL 13: 3 CR 9 Guardians with a CR 7 hazard and 2 CR 5 traps
EL 13: 2 sets of (4 CR 6 Swarmlings with a CR 8 Commander)
EL 13: 2 sets of (8 CR 4 Swarmlings with a CR 6 Squad Leader) with 1 CR 8 Commander.
EL 13: An ambush with 2 CR 7 Hosers 2 CR 7 Artillery and 4 CR 5 Assassins.
PS: These are all TPKs in 5E. Don't try them.

Edit: If only we could combine that 3E ease with simple stat blocks like 5E.

Are you confusing 3E with 4E?

Because in 3E and its derivatives, I could make a party that could curbstomp those EL 13 fights, or one that could get bopped so hard by them. Power is wildly inconsistent across the same level in 3.5, depending on builds.

[Selion]

Depends. They kinda can but it's a bit of a different ballgame at that point. Well, playing Tarrasque would probably not be very gratifying except as a one-shot but a game of Ancient Dragons could make for a very engaging Tier 4ish campaign.

RP an ancient dragon, sure!
"master, i think I'll sleep on my hoard for a decade, wake me up if the world is ending or something..."

[Tanarii]

Yeah, 3E really spoiled me by letting the PCs power curve and the monster power curve be smooth exponentials of roughly the same base.

Thats nothing like the 3e I played. Different classes had different power curves. They weren't even necessarily curves, they could be a jagged line. And Multiclassing/PrCs made a joke of the entire concept of a power curve of any kind.

[OldTrees1]

But 5E is so easy by default (if you avoid trick monsters and advanced tactics) that I can still do the same thing anyway in 30-45 seconds. Level 13 fight: how about one CR 13 Devourer, a couple of CR 5 Star Spawn Manglers, and let's add three CR 4 Flameskulls as magical support. Same 45 seconds to compute, similar level of certainty (it won't be a TPK unless the party is exceptionally weak but will be strong enough to feel dangerous/punish mistakes). Same flaws too: have to know the monster abilities, not just the CR, and lots of variance in the actual power level of said level 13 party based on builds and tactics.

CR just isn't very precise (the difference in stats between CR 6 and 8 is small, just look at white vs. green young dragons), so any variance from guesstimating difficulty is swamped by the variance in party builds/tactics anyway as well as dice variance.

I just double checked that encounter you described using an online calculator that uses the 5E DMG rules. The same rules 5E thought to give new or transitioning players / DMs. I did not bother spending time doing the math by hand. I see that 45 second design results in a EL of 21 and more than 3x the Deadly threshold. If you are right and that is a fine encounter, the DMG does not communicate that to me. Double checking, my EL 13 examples were because the party was level 10. If you meant level 13 then it is only 1.5x the Deadly threshold. Still, I don't see 5E as succeeding on this front.

Are you confusing 3E with 4E?

Because in 3E and its derivatives, I could make a party that could curbstomp those EL 13 fights, or one that could get bopped so hard by them. Power is wildly inconsistent across the same level in 3.5, depending on builds.

You missed context upthread in the monster stat block subthread. I am using CR as the actual challenge relative to the party, not the WotC estimate. So that would factor in the party optimization level.

This is another reason why I like 5E simplified monster/NPC generation rules.

[Tanarii]

and more than 3x the Deadly threshold. If you are right and that is a fine encounter, the DMG does not communicate that to me.

Max has some heavily divergent ideas from the DMG, and for that matter from what I've seen most DMs on these forums, about what constitutes a "fine encounter". His bare minimum seems to be 2xDeadly.

Edit: but I should be clear, he's very consistently divergent.

[MaxWilson]

I just double checked that encounter you described using an online calculator that uses the 5E DMG rules. The same rules 5E thought to give new or transitioning players / DMs. I did not bother spending time doing the math by hand. I see that 45 second design results in a EL of 21 and more than 3x the Deadly threshold. If you are right and that is a fine encounter, the DMG does not communicate that to me. Double checking, my EL 13 examples were because the party was level 10. If you meant level 13 then it is only 1.5x the Deadly threshold. Still, I don't see 5E as succeeding on this front.

It's actually 33,800 adjusted XP, about 50% over the Deadly threshold of 20,400 for four 13th level PCs, which is pretty much what I was aiming for: nontrivial (there's a chance of someone dying = Deadly by definition) but not too hard. If you got 3x Deadly there's a bug in the calculator you used. My experience teaches me that this fight is not too tough for 13th level PCs--I invite you to run a test fight and you'll see. (Try to take out the manglers ASAP because they are glass cannons.)

By other's accounts it sounds like 3E was about the same as 5E in this regard, maybe worse, due to more variability in PC power.

Max has some heavily divergent ideas from the DMG, and for that matter from what I've seen most DMs on these forums, about what constitutes a "fine encounter". His bare minimum seems to be 2xDeadly.

Edit: but I should be clear, he's very consistently divergent.

In this case it's not even 2x Deadly. Note that WotC has published encounters of a similar difficulty, e.g. Rime of the Frost Maiden is full of Deadly+ fights.

For the record I don't mind having trivially-easy encounters in my adventures, I just don't like spending table time on them. If the party splits up to investigate rumors and a lone 13th level PC runs across a dozen bandits led by an Oni and the situation degenerates into combat, I'll play that fight out (it turns out to be Deadly x2 BTW) but if the whole party is there I'd rather just skip to the next part--there's no dramatic question about whether the PCs can win that (Medium) curbstomp. I don't like dragging scenes out after the dramatic question is resolved.

[OldTrees1]

It's actually 33,800 adjusted XP, about 50% over the Deadly threshold of 20,400 for four 13th level PCs, which is pretty much what I was aiming for: nontrivial (there's a chance of someone dying = Deadly by definition) but not too hard. If you got 3x Deadly there's a bug in the calculator you used. My experience teaches me that this fight is not too tough for 13th level PCs--I invite you to run a test fight and you'll see. (Try to take out the manglers ASAP because they are glass cannons.)

By other's accounts it sounds like 3E was about the same as 5E in this regard, maybe worse, due to more variability in PC power.

The 3x was when I checked the same 10th level party (EL 13 fight for a 10th level party was a boss fight in 3E). I noticed part way through that you upgraded them to 13th.

I barely doubt you about whether the fight actually is / is not too tough. However the DMG math (add up the xp values and use a flat multiplier based on number of monsters) does not give me the exponential power curve that makes it easy to go:

Spoiler: Quick example

Show

Start
13
12 10
10 10 10
10 8 8 4 4 4 4
Done

So 5E math has me do guess and check. And after the guess and check I am left with a wrong answer. But despite the answer being wrong, I am still building the intuitions to correct it. Basically 5E made it hard to use and hard to trust the encounter design rules. Or at least that is my personal experience (which is dominated by mixed CRs and unequal ratios).

Tanarii's experience with 3E differs from my curated experience. Our group maintained a consistent optimization level that did not see anything nearly as jagged as 5E tiers levels. The kind of power curves I saw were what the 3E DMG expected, as long as I estimated the CR of monsters myself based on our consistent optimization level.

[MaxWilson]

The 3x was when I checked the same 10th level party (EL 13 fight for a 10th level party was a boss fight in 3E). I noticed part way through that you upgraded them to 13th.

Oh, I had no idea you wanted 10th level. I just gave you a "bossfight" (i.e. minor adventure climax) off the top of my head, aimed at a level 13 party. I guess I misunderstood what you were trying to convey when you said "EL 13."

For a level 10 party, call it an Abominable Yeti and two Wraiths plus a Bone Naga.

I barely doubt you about whether the fight actually is / is not too tough. However the DMG math (add up the xp values and use a flat multiplier based on number of monsters) does not give me the exponential power curve that makes it easy to go:

Spoiler: Quick example

Show

Start
13
12 10
10 10 10
10 8 8 4 4 4 4
Done

So 5E math has me do guess and check. And after the guess and check I am left with a wrong answer.

5E power curves are closer to linear than exponential, except at CRs under 1. When I built that 13th level fight in my head I was basically equating CR = Level (4 CR 13 creatures), then deducting one monster to tilt the odds in the party's favor (3 CR 13s), then converting some of them into weaker creatures (1 CR 13, two CR 5s, three CR 4s) to tilt it even more in the party's favor while giving more variety to the adventure. Ditto for the level 10 fight.

5E power curves aren't really linear because the slope of the power curve changes at certain points (IIRC around CR 10 it gets a little steeper and then at least 20ish the slope gets three times steeper) but it's definitely much closer to linear than exponential. Instead of telling yourself that CR 10 = 8+8, tell yourself that CR 10=5+5ish.

[OldTrees1]

Oh, I had no idea. I just gave you a "bossfight" (i.e. minor adventure climax) off the top of my head, aimed at a level 13 party. I guess I misunderstood what you were trying to convey when you said "EL 13."

That is my bad. However you labeled your example well enough that I eventually noticed and recalculated the x1.5 threshold.

5E power curves are closer to linear than exponential, except at CRs under 1. When I built that fight in my head I was basically equating CR = Level (4 CR 13 creatures), then deducting one monster to tilt the odds in the party's favor (3 CR 13s), then converting some of them into weaker creatures (1 CR 13, two CR 5s, three CR 4s) to tilt it even more in the party's favor while giving more variety to the adventure.

5E power curves aren't really linear because the slope of the power curve changes at certain points (IIRC around CR 10 it gets a little steeper and then at least 20ish the slope gets three times steeper) but it's definitely much closer to linear than exponential. Instead of telling yourself that CR 10 = 8+8, tell yourself that CR 10=6+5ish.

Linear? As in: (CR,Power) follows F(X) = mx + b? That sounds too good to be true.
2 CR X is the same power as 1 CR X+a and 1 CR X-a?
Oh and you are saying "b" is small so 2 CR X might be the same power as 1 CR 2X?

This really sounds too good to be true. If this is true then this should have been in the 5E DMG. That would have made things much easier. I apologize for being so skeptical, but this does sound too good to be true. Linear and exponential can both make really fast design math.

[MaxWilson]

Linear? As in: Power(CR) = m(CR) + b? That sounds too good to be true.
2 CR X is the same power as 1 CR X+a and 1 CR X-a?
Oh and you are saying "b" is small so 2 CR X might be the same power as 1 CR 2X?

This really sounds too good to be true. If this is true then this should have been in the 5E DMG. That would have made things much easier. I apologize for being so skeptical.

It's not something the DMG says outright, it's something you notice when you examine the table on DMG page 274 closely. Between CR 1 and CR 19, every +1 to CR adds 15 HP and 6 damage. It's also something you can derive from the encounter construction rules (e.g. compare adjusted XP two CR 3s to a CR 6). The AC and to-hit progressions are not as neat but still basically linear.

The relationship is not neat and it breaks down in several places (a CR 30 monster has the stats you'd predict for a CR 50ish monster if you extrapolated from CR 1 to 19, and of course CR 1/8 to 1 has a very strange power progression indeed). But it's a smoother progression than the PC power progression and since 5E is meant to be tilted in favor of the players anyway, it's easy to guesstimate a fight by erring on the side of the PCs.

It's not like anyone has an exact, objective metric anyway for determining if a fight was the "right difficulty." The 5E designers got their table from playtesting AFAICT, just fitting a curve to whatever level of difficulty made their playtesters happy in the dungeon crawls they were using during the playtest, and I'm sure that the CRs over 20 got minimal playtesting anyway. There's no secret master plan.