What levels are NPCs, and how common is spellcasting? [Archive]

Greywander

2018-07-02, 03:16 AM

The general impression I have is that the majority of NPCs would be fairly low level if modeled with class levels instead of just a stat block. An example of this is the PHB description of clerics, where it specifically calls out that not all priests would have cleric levels. However, looking at the NPC entries in the Monster Manual doesn't seem to bear this out. Perhaps most poignant is the Acolyte NPC, who is a 1st level spellcaster, despite what the PHB says about clerics.

There's some other weird examples, too. For example, one would think that we could extract an NPC's level from their hit dice, but this seems to be inconsistent. The previously mentioned Acolyte is specifically called a 1st level spellcaster, but has two hit dice. The Archmage has 18 hit dice and is an 18th level spellcaster. Then there's the Assassin (no I haven't gotten past the "A"s yet), which has 12 hit dice, but only does 4d6 Sneak Attack damage, suggesting a rogue 7 or 8. Maybe the Assassin is multiclassed, but into what? He doesn't even have all of the rogue abilities he should have, let alone the features of any other class.

Moving further into the list of NPCs, we see that the Bandit Captain has 10 hit dice. The Cult Fanatic has 6 hit dice and is a 4th level spellcaster. The Druid has 5 hit dice and is a 4th level caster, while the Mage has 9 hit dice and is a 9th level caster (I guess wizards don't multiclass, at least not if they're NPCs). The Guard has a pathetic two hit dice, the Knight a more impressive 8, and the Gladiator a massive 15 hit dice. Something unusual about the Gladiator is that despite having 15 hit dice, his proficiency bonus is only +3, placing him between levels 5 and 8.

I really don't understand what's going on here, or what logic is being used to determine NPC statistics. It would make the most sense to me to model NPC the same way we do PC, which would make it easier to compare their power levels and make sure they pose an appropriate challenge (either when used as antagonists, or if the PCs decide to attack them).

(Upon further inspection, I've noticed that all NPCs seem to use a d8 for their hit dice, which might be why e.g. the Gladiator seems to have far more hit dice than his proficiency bonus suggests. And unlike PCs, they don't get to "max out" their first hit die, i.e. a 1st level cleric starts with 1d8 hit dice and 8 HP, while an Acolyte NPC has 2d8 hit dice and 9 HP.)

How common magic is also has a big impact on the setting the game takes place in. For example, what originally got me looking into this was wondering what sort of magical support a city guard might have. This will also affect how NPCs react to PC spellcasting, and how hard it is to find an NPC spellcaster (healers, magic item shop, etc.) whose services the PCs can hire.

These two aspects (level and spellcasting) are intertwined, so if we can unravel one side it might help unravel the other. For some reason, I've gotten it into my head that spells up to 5th level are generally known, but 6th level and up are "secret" or "lost" or otherwise outside of public knowledge. Rumors might exist of a powerful wizard or cleric living in the capital city or on top of a remote mountain, but you could probably count folks like this on one hand. Since spells get a pretty big bump at 6th level, and that's also a convenient cut-off point for resurrection magic (Raise Dead is common knowledge, if not so common to find someone able to cast, while Resurrection and True Resurrection are mere rumors and legends).

Going by this, we can develop a rough framework for what level an NPC should be, and what that says about their level of competence in the world:

1st level - Trained
Even gaining one class level is a feat in and of itself. You're no novice, even if you're not an expert yet.

5th level - Professional
By this point, you've chosen a subclass and gained some of your class's core features. You're a pro.

11th level - Master
You are very likely stronger than the people that originally trained you. You are, as far as you know, at the peak of your ability.

17th level - Legend
Your abilities have grown to mythical proportions, and you're capable of feats spoken of only in epic poems that none thought truly possible.

If we want a finer grain, we can use proficiency bonus instead of tiers.

1-4th level (+2 prof.) - Trained
5-8th level (+3 prof.) - Professional
9-12th level (+4 prof.) - Master
13-16th level (+5 prof.) - Heroic
17-20th level (+6 prof.) - Mythical

This might actually be better since it helps us gauge what an NPC's proficiency bonus should be.

An observation from this is that we might expect the professors at a wizard school to be capable of casting spells from 3rd to 6th level (Professional/Master range). If it is a prestigious institution, there might be one or a few wizards capable of more powerful spells. A typical thieves' guild member is likely to fall in the Professional range (with a few grunts/lackeys in the Trained range), while the guild master might fall in the Master range, but could also be at the high end of the Professional range.

"Master" generally represents the highest level a typical NPC will ever achieve in their lifetime, and then only through much hard work. These people are almost always noteworthy and tend to occupy prestigious positions in society. Levels above Master transcend human limits and are probably going to be reserved for "major" NPCs instead of random NPCs. For example, the patron who sends you out on missions might be a higher level, or the main villain, or basically any recurring NPC. But the shopkeeper? The caravaner? The guardsman at the city gate? Nah, those will be Trained or Professional at best.

Well this turned into kind of a long, rambly post. I guess this means I should stop typing and get some sleep.

Platypusbill

2018-07-02, 03:56 AM

NPCs just tend to be built differently than PCs.

-They don't have any class levels, even though they sometimes have features of PC classes, such as Sneak Attack.

-Level 1 PCs use the maximum amount of their hit die for their level 1 HP (e.g. a level 2 Cleric has 8 + 1d8 HP before Constitution and other modifiers). Meanwhile, the HP pools of an NPC is entirely rolled, e.g. 2d8.

-NPCs always use d8s for their hit dice because they are Medium creatures by default (Small creatures use d6s, Large creatures use d10s). This is in contrast to PCs, who use d6-d12 Hit Dice depending on their class. The amount of hit dice is set to an arbitrary amount so the NPC reaches a desired HP pool.

-Generally, NPCs and monsters have huge HP pools compared to PCs, because otherwise they would be quickly overwhelmed in a situation where there are one or two powerful monsters and a bunch of mooks against a party of PCs. A CR5 monster is designed to put up a fight against several level 5 PCs at once.

Unoriginal

2018-07-02, 04:35 AM

The general impression I have is that the majority of NPCs would be fairly low level if modeled with class levels instead of just a stat block.

NPCs don't have levels, but generally you can say that a level is between 1/4 and 1/3 of a level.

An example of this is the PHB description of clerics, where it specifically calls out that not all priests would have cleric levels. However, looking at the NPC entries in the Monster Manual doesn't seem to bear this out. Perhaps most poignant is the Acolyte NPC, who is a 1st level spellcaster, despite what the PHB says about clerics.

MOST priests aren't Acolytes. The Acolyte NPC statblock is for the minority who indeed are Clerics.

Not every NPC with the function of religious acolyte uses an Acolyte NPC statblock.

There's some other weird examples, too. For example, one would think that we could extract an NPC's level from their hit dice, but this seems to be inconsistent.

It's acknowledged and done on purpose that NPC hit dices have no correlation to classes. All Medium humanoids have d8, for example.

I really don't understand what's going on here, or what logic is being used to determine NPC statistics.

The logic is "NPCs are not built like PCs", "NPCs are easier to make and adjust if they're not tied to PC classes" and "NPCs are tailored for their role in the world, not by PC classes.

It would make the most sense to me to model NPC the same way we do PC

No.

This is antithetic to the very basis of 5e design.

They tried that before and it just made creating NPCs a mess.

which would make it easier to compare their power levels and make sure they pose an appropriate challenge (either when used as antagonists, or if the PCs decide to attack them).

No, it would not. The CR system already does that.

(Upon further inspection, I've noticed that all NPCs seem to use a d8 for their hit dice, which might be why e.g. the Gladiator seems to have far more hit dice than his proficiency bonus suggests. And unlike PCs, they don't get to "max out" their first hit die, i.e. a 1st level cleric starts with 1d8 hit dice and 8 HP, while an Acolyte NPC has 2d8 hit dice and 9 HP.)

Yes, that is acknowledged and done on purpose.

Also, the proficiency bonus is based on the CR, which is why the Gladiator has +3.

How common magic is also has a big impact on the setting the game takes place in. For example, what originally got me looking into this was wondering what sort of magical support a city guard might have. This will also affect how NPCs react to PC spellcasting, and how hard it is to find an NPC spellcaster (healers, magic item shop, etc.) whose services the PCs can hire.

That is discussed in the PHB and DMG, as well as the specific module you're using.

These two aspects (level and spellcasting) are intertwined, so if we can unravel one side

NPC statblock precise "X is a lvl Y spellcaster", when they're spellcasters. There is literally nothing to unravel.

For some reason, I've gotten it into my head that spells up to 5th level are generally known, but 6th level and up are "secret" or "lost" or otherwise outside of public knowledge.

I don't know why you got this into your head, but it's certainly not an assumption of the game.

That being said, it's true many spell casters aren't good enough to cast spell above the 5th level. It doesn't mean lower levels are generally known, or that higher are secret or whatever.

Rumors might exist of a powerful wizard or cleric living in the capital city or on top of a remote mountain, but you could probably count folks like this on one hand. Since spells get a pretty big bump at 6th level, and that's also a convenient cut-off point for resurrection magic (Raise Dead is common knowledge, if not so common to find someone able to cast, while Resurrection and True Resurrection are mere rumors and legends).

Going by this

You can't start with an unproven, unsupported assumption and then create a system around it.

1st level - Trained

5th level - Professional

11th level - Master

17th level - Legend

Dude, now you're trying to reinvent the tiers of plays.

NPCs aren't PCs.

If we want a finer grain, we can use proficiency bonus instead of tiers.
[...]
This might actually be better since it helps us gauge what an NPC's proficiency bonus should be.

NPC proficiencies are determined by their CR, or by their work. The dinosaur trainers of Chult have +5 in Wis(Animal Handling), despite being Commoners.

You DO NOT get to pretend knowing what they "should be".

An observation from this is that we might expect the professors at a wizard school to be capable of casting spells from 3rd to 6th level (Professional/Master range). [QUOTE=Greywander;23192574]

Or you could look at the Mage statblock, or at the various specialist wizard statblocks in the Volo's.

[QUOTE=Greywander;23192574]A typical thieves' guild member is likely to fall in the Professional range (with a few grunts/lackeys in the Trained range), while the guild master might fall in the Master range, but could also be at the high end of the Professional range.[QUOTE=Greywander;23192574]

Or not. You could also look at the

[QUOTE=Greywander;23192574]
"Master" generally represents the highest level a typical NPC will ever achieve in their lifetime, and then only through much hard work. These people are almost always noteworthy and tend to occupy prestigious positions in society. Levels above Master transcend human limits and are probably going to be reserved for "major" NPCs instead of random NPCs. For example, the patron who sends you out on missions might be a higher level, or the main villain, or basically any recurring NPC. But the shopkeeper? The caravaner? The guardsman at the city gate? Nah, those will be Trained or Professional at best.

Or they'll be whatever.

A king could have the Noble NPC statblock with a +7 in Wis (Insight), or a Knight statblock, or an Assassin statblock. A guard could be a Guard, a Bandit Captain, a Veteran, or whatever.

In Tomb of Annihilation, the Merchant Prince in control of all the magic item business is a Mage, and so can cast spells up to 5th lvl. Another Merchant Prince, the most powerful human priestess on the island is a Priest, and so can cast spells up to 3rd lvl. Other Merchant Princes include a Gladiator, an Assassin, a Scout and two Nobles.

They have on their payroll a Sorcerer, whose statblock indicates being a lvl 14 spellcaster who can cast spells up to the 7th lvl. This guy is the harbor master, aka the person in charge of making sure the boats are orderly and that no one smuggle stuff.

I have no idea what you're trying to say except "guys, powerful people are rare and you shouldn't have random NPCs be that good"?

JackPhoenix

2018-07-02, 06:52 AM

There's a reason why "NPCs use the same rules as PCs" was abandoned after being used in only a single edition, which was an overcomplicated mess overall.

Glorthindel

2018-07-02, 09:12 AM

Xetheral

2018-07-02, 06:31 PM

Personally I just ignore the NPC statblocks in the MM and just build like a PC everyone that I need stats for (albeit not always in full detail).

Building NPC's and monsters using different systems, and in some places by applying slightly different rules for the same mechanism (the differences between multiattack and extra attack is just inanely stupid) is just bewildering to me. I have always found that a game works better (and feels fairer) when everyone is playing by the same rules.

Agreed. I treat as fundamental principle that similar things should be modeled (or abstracted) similarly. I prefer to run games where there isn't any difference between PCs and NPCs in the game world, and mechanically modeling then consistently helps reinforce that impression.

Consistent modeling of PCs and NPCs also means players can draw reasonable inferences about an NPCs abilities from at-the-table oberservations. That makes NPCs relatable, rather than seeming like the random conglomeration of stats and abilities found in the monster manual. I find it helps immersion when the NPCs seem less abstract.

Also, I entirely agree with you about the ultra-technical distinction between Extra Attack and Multiattack being utterly inane.

PhoenixPhyre

2018-07-02, 07:35 PM

You can stat all humanoid NPCs as PCs, but to me it comes at a tremendous cost.

1) It's substantially more effort. And no, pre-filled templates don't take all of that away. So you end up using fewer NPCs and/or more monsters (who don't use NPC classes at all unless you want to do serious adaptation).
2) It's sharply limiting--you are restricted to the set of things that are balanced for long-term PC use, and have to give them the appropriate levels to do so, increasing their hit die and combat capabilities accordingly. Thus, you can't have a non-combatant priest. They all have medium/heavy armor and a crap ton of damage abilities. You're restricted to having all your important NPCs as high level people, otherwise they can't have appropriate abilities. Then you run into the same power-creep that makes FR (and 3e settings in general) so incoherent.
3) It reifies classes and levels as being actual, in-universe things, with all the attendant weirdness that promotes.
4) It means you throw out all the balancing guidelines entirely, unless you also want to go and recalculate CR for each of these NPCs (which will not follow the same guidelines, since NPCs and PCs are designed to be different). CR =/= level (not even close). CR =/= HD (not even close).

NPCs (and monsters) and PCs are asymmetric in role. PCs face lots of NPCs each day, while NPCs only face one set of PCs, often in their lifetime (on screen anyway). NPCs (and monsters, the two are equivalent in role) should come in large numbers (>> 1). This means that while PCs can be complex (since each player only controls one), NPCs have to be simple (otherwise the DM can't run them all or resorts to fewer, smaller-number, more powerful fights with all the pathologies that entails). So most of the work that goes into creating this "PC-NPC" is wasted--it never shows up on screen or has an in-game effect.

I've actually gone the other way with spell-casters--I choose 3-4 spells (max) that they'll cast during combat and the rest are as needed based on the narrative and the personality. I also don't restrict myself to the PC-class spell lists. Those are game artifacts for balance purposes. I also explicitly state that there are thousands of spells and abilities out there that aren't in the PHB or any printed book. Some people might be the only ones who can do that, because they're special (or have the appropriate knowledge, connections, and bloodline).

Heck, I have an NPC wandering around who has the following trait:

Living Library: Iulia can prepare and cast any spell of 6th level or lower as if it were a wizard spell, obeying the usual restrictions and limits on number of spells prepared. She does not require a spell-book.

That would be seriously out-of-bounds for a PC, but as DM (and world-builder) I can guarantee that Iulia won't mess anything up. Likely, she'll never show up in another game--she was rescued by a set of PCs and now is wandering the world, seeing new sights and randomly helping people.

I can freely improvise rituals, knowing that they're bounded by my decisions and not vulnerable to munchkin gaming.

All in all, I'm glad that 5th edition abandoned PC-NPC transparency. It was the root cause of so much of 3e's dysfunction, and actively weakens the game in my opinion.

Armored Walrus

2018-07-02, 08:12 PM

I don't think you can define what "most" priests are based on any stat block in the MM. The stuff in the MM is stuff that the PCs are meant to fight. "Acolyte" is simply the name of that stat-block. In-world, I think "most" acolytes would run screaming if four armed and armored individuals busted through the church doors. (remember acolytes are just kids)

ZorroGames

2018-07-02, 08:20 PM

Spell casting is as common as the DM wants it to be.

NPCs =/= PCs.

JoeJ

2018-07-02, 08:26 PM

MrStabby

2018-07-02, 08:37 PM

I find that spellcasting has to be relatively common at least to preserve some balance in the game.

In a game where the threats are all non-magical then the defences are all against non magical attacks. Locked doors secure treasure, watchmen look out for infiltrators and armour protects you from harm. This then gives casters disproportionate power with spells like invisibility, knock, hold person and so on. If on the other hand magic is prevalent in the world then opposing casters are likely to have counterspell prepared as well as Nystul's Aura and absorb elements. Rings of resistance will be prized not be cheap trinkets. Castles will have more than wall for protection and are likely to have tricks that hinder those who use the fly spell or spiderclimb. Soldiers will be trained to spread out and approach from different sides rather than cluster in fireball sized blobs. Downed enemies will be finished off in the knowledge that healing magic exists.

The balance between caster and martial is contingent upon the NPCs in the game world knowing they have to defend against both, not just one of these threats.

Armored Walrus

2018-07-02, 08:57 PM

The "PC's and NPC's are made by different rules" is possibly my biggest (and maybe only) gripe about 5th edition.

Through all previous editions of D&D (barring 4th), and many other systems, I have always built NPCs using the same rules as player characters. Contrary to the prevailing claim made by people who likely haven't done it, it is not difficult, and its not even particularly time consuming - I did it easily enough back when I hand-wrote all my DM notes, it is not going to get more difficult, or take longer these days with computers and reuseable templates that auto-fill half the data if you set it up right.

Building NPC's and monsters using different systems, and in some places by applying slightly different rules for the same mechanism (the differences between multiattack and extra attack is just inanely stupid) is just bewildering to me. I have always found that a game works better (and feels fairer) when everyone is playing by the same rules.

To be fair, it is one reason I am glad for Volo's guide - I may not be a fan of the lore sections, but it gives me the info I need to create key boss monsters in the manner I prefer.

You may have built NPCs using the same rules as PCs, but the monster manuals of previous editions didn't necessarily do so. AD&D 2e Monstrous Manual, for instance: Entries have Climate/Terrain, Frequency, Oganization, Activity Cycle, Diet, Intelligence, Treasure, Alignment, No. Appearing, Armor Class, Movement, Hit Dice, THAC0, Number of Attacks, Damage per attack, Special Attacks, Special Defenses, Magic Resistance, Size, Morale, XP Value. No STR, DEX, CON, WIS or CHA, so no CON modifier, etc.

So while you could fully create your NPCs from PC rules, not every edition assumed you would, and I think there's more history of "the stuff you fight" using different rules than PCs than there is of them using the same rules. And if your position is "yeah, but the PC's aren't necessarily fighting NPCs; NPCs aren't monsters" then I say to you, why do you need stats if they aren't going to fight it? Just tell the story of what the NPCs do and leave it at that.

Tanarii

2018-07-02, 09:13 PM

4) It means you throw out all the balancing guidelines entirely, unless you also want to go and recalculate CR for each of these NPCs (which will not follow the same guidelines, since NPCs and PCs are designed to be different). CR =/= level (not even close). CR =/= HD (not even close).

NPCs (and monsters) and PCs are asymmetric in role. PCs face lots of NPCs each day, while NPCs only face one set of PCs, often in their lifetime (on screen anyway). NPCs (and monsters, the two are equivalent in role) should come in large numbers (>> 1). This means that while PCs can be complex (since each player only controls one), NPCs have to be simple (otherwise the DM can't run them all or resorts to fewer, smaller-number, more powerful fights with all the pathologies that entails). So most of the work that goes into creating this "PC-NPC" is wasted--it never shows up on screen or has an in-game effect.This is the main reason when I am DMing or playing, I view NPCs built as PCs as a bad thing. Much more so when DMing though.

I want a good way to judge how dangerous they are (roughly) and I want my DM to have a good idea how dangerous they are (roughly). The DM can, of course, calculate that using the DMG guidelines. But how many spend time doing that? I certainly don't have time for that. This applies when I DM, but also when playing.

I also don't want an NPC balanced on the PC paradigm of many encounters with monsters per rest, instead of one encounter with the PCs.

I also don't want NPCs that are complex as all get-out to run at the table. Players only have one PC, or maybe one PC plus one henchman to run. I have many monsters to run.

From what I've seen, DMs want NPCs to be their own little PCs, not another monster. I don't want that. To me, NPCs and Monsters are synonymous. They should run using the same rules by default. Don't make me use two different systems to run a game. (Edit: and no, making monsters use PC rules is not the solution to this. The PC rules are too complex for a DM to use. Again, I have many monsters to run.)

Of course, by default is the key word. An occasional exception is fine. As the DMG notes, NPCs can be made using PC rules if you really want to.

I have a specific type of them in my campaigns. They're called "henchmen" and fight on the side of the PCs. But even that's not required. Many AL adventures, and the hardcover adventure paths, stat out "monster" stat blocks for NPCs even though they will be on the side of the players, and even potentially under the control of the players. Personally I like henchmen and player controlled allies as PCs better for the same reason I like NPCs and DM controlled creatures as monsters. Don't make them learn to use two different systems to play the game.

Pex

2018-07-02, 10:19 PM

Their level and spellcasting ability is as common or varied as the DM wants them to be for the gameworld. There is no set rule on this. At best a published reference guide for a ready made gameworld could provide population data for the DM to use, but this is still all about world building and of the complete purview of the DM. With different DMs having different populations of NPCs and their spellcasting ability, neither is doing it wrong.

If you're a DM looking for advice on how different population numbers could determine atmosphere of the gameworld, that's worth discussing. If you're a Player, the only thing that matters is the gameworld you're playing in, but it doesn't hurt to understand the different atmospheres as well. There is no definitive correct answer on what should be.

Tanarii

2018-07-02, 11:07 PM

There is no set rule on this.
Clearly the correct balance is 1/100 classed characters in dangerous frontier areas, and 1/1000 in civilized areas. With 1/2 level 1, and 1/2 of the remainder level 2, and so on.

I can't even remember where I got this scheme from any more, but it's my basic assumption for "correct" D&D population distribution now. Even when it doesn't fit with the assumptions of an edition or campaign setting. :smallamused:

JoeJ

2018-07-02, 11:21 PM

Clearly the correct balance is 1/100 classed characters in dangerous frontier areas, and 1/1000 in civilized areas. With 1/2 level 1, and 1/2 of the remainder level 2, and so on.

I can't even remember where I got this scheme from any more, but it's my basic assumption for "correct" D&D population distribution now. Even when it doesn't fit with the assumptions of an edition or campaign setting. :smallamused:

According to the original AD&D DMG: "Human and half-orc characters suitable for level advancement are found at a ration of 1 in 100. Other races have an incidence of 1 in 50."

As I recall, 3.x has tables to determine the class & level of everybody in a town. For 5e, it's just whatever the DM decides. (Which is as it always should have been IMO.)

Vessyra

2018-07-02, 11:37 PM

After switching from 3.5, I was confused by how the NPCs didn't have class levels. I've since realised that NPCs don't use PC levels, which I'm glad of. It makes balancing easier, since you don't have to worry about what kind of change that ability X would do to the CR. Plus I think it makes PCs more special; there's not much fun in your cleric walking into the city and discovering that there are several dozen more clerics with the exact same abilities as him walking around.

With the 1st level acolyte, I think that it was meant to be second level, not because they have two hit dice but because they have three spell slots

Xetheral

2018-07-03, 12:13 AM

You can stat all humanoid NPCs as PCs, but to me it comes at a tremendous cost.

1) It's substantially more effort. And no, pre-filled templates don't take all of that away. So you end up using fewer NPCs and/or more monsters (who don't use NPC classes at all unless you want to do serious adaptation).
2) It's sharply limiting--you are restricted to the set of things that are balanced for long-term PC use, and have to give them the appropriate levels to do so, increasing their hit die and combat capabilities accordingly. Thus, you can't have a non-combatant priest. They all have medium/heavy armor and a crap ton of damage abilities. You're restricted to having all your important NPCs as high level people, otherwise they can't have appropriate abilities. Then you run into the same power-creep that makes FR (and 3e settings in general) so incoherent.
3) It reifies classes and levels as being actual, in-universe things, with all the attendant weirdness that promotes.
4) It means you throw out all the balancing guidelines entirely, unless you also want to go and recalculate CR for each of these NPCs (which will not follow the same guidelines, since NPCs and PCs are designed to be different). CR =/= level (not even close). CR =/= HD (not even close).

It isn't THAT much more effort to stat an NPC as a PC instead of making a custom statblock, particularly if you don't bother making the character in full detail. Heck, in some ways it's faster: most of the stats for PC characters are derived from the build, whereas when making an NPC as a statblock there are a lot more arbitrary numbers that require a decision. Restricting high-level abilities to high-level characters I consider a benefit, not a cost. It makes the game world more dependable and less random. You can still keep classes as purely out-of-game concepts, even when restricting NPCs to PC rules. For this one you make a very good point, and one I often forget. I don't use CR at my table anyway, so it doesn't matter to me, but to DMs who want to use CR it's definitely harder to build NPCs as PCs.

This is the main reason when I am DMing or playing, I view NPCs built as PCs as a bad thing. Much more so when DMing though.

I want a good way to judge how dangerous they are (roughly) and I want my DM to have a good idea how dangerous they are (roughly). The DM can, of course, calculate that using the DMG guidelines. But how many spend time doing that? I certainly don't have time for that. This applies when I DM, but also when playing.

Out of curiosity, why is CR important to you? You're said you run a combat-as-war sandbox. To me that would suggest that there is no expectation that encounters are balanced, making CR somewhat irrelevant. By contrast, in a CaW sandbox it's important to telegraph combat difficulty, and it's a lot easier to do that in-game when NPCs are built on PC rules: just have the NPC use an ability in sight of the PCs, and the minimum level of that NPC is immediately obvious. If NPCs are just statblocks, the PCs can't infer anything about the opponent's difficulty from in-game observations without a lot more data points.

Tanarii

2018-07-03, 12:22 AM

According to the original AD&D DMG: "Human and half-orc characters suitable for level advancement are found at a ration of 1 in 100. Other races have an incidence of 1 in 50."

As I recall, 3.x has tables to determine the class & level of everybody in a town. For 5e, it's just whatever the DM decides. (Which is as it always should have been IMO.)
AD&D was a different mindset, really designed for gaming organizations with many different players, or tourney play. Plus, y'know, partially designed by an actuarial-minded guy who loved tables, stats, and had a belief that being a fair minded referee/judge meant things generate randomly.

That said, some world building guidelines as suggestions aren't a terrible thing for newer DMs, but following them slavishly isn't a great idea. I fondly remember the 2e World Builder's Guidebook. But all it ever did was give me ideas, I never used it to randomly generate a continent or kingdom and populate it. :smallamused:

Edit: it's worth noting that the AD&D population stats you quoted would be important for finding henchmen. Is that the section it came from?

Edit2
Out of curiosity, why is CR important to you? You're said you run a combat-as-war sandbox. To me that would suggest that there is no expectation that encounters are balanced, making CR somewhat irrelevant. By contrast, in a CaW sandbox it's important to telegraph combat difficulty, and it's a lot easier to do that in-game when NPCs are built on PC rules: just have the NPC use an ability in sight of the PCs, and the minimum level of that NPC is immediately obvious. If NPCs are just statblocks, the PCs can't infer anything about the opponent's difficulty from in-game observations without a lot more data points.Knowing how dangerous a thing is for a group is just as important for a CaW game for exactly the reason you noted: to know how to telegraph potential danger. I don't design encounters for a specific party, but I do generally design adventuring areas for a rough difficulty. And then telegraph its danger to the party based on their relative party level.

If I didn't know the difficulty, I couldn't telegraph it, and players would wipe even more often than they already do, because they'd be acting blind without any information.

JoeJ

2018-07-03, 12:53 AM

Edit: it's worth noting that the AD&D population stats you quoted would be important for finding henchmen. Is that the section it came from?

That is indeed where it came from. It also mentions that most NPCs with levels are already in jobs they're happy with, so only about 1 in 1000 population are available as henchmen. Depending on how wild the area is this can be adjusted as high as 1 in 200 or as low as 1 in 5000.

Regarding rules for creating NPCs; in the largest city of my world there is a brother & sister pair of thieves whose street names are Lock and Key. The girl (Lock) can cast Arcane Lock at will. The boy (Key) can cast Knock at will. In every other respect they're just moderately skilled thieves, with no other magical abilities. I give important NPCs whatever abilities I think are interesting and fit their role in the world. A fencing instructor, for example, might apply double her proficiency bonus to attacks with her preferred weapon, or a sage might be able to cast every ritual spell but not any spells otherwise. If I had to restrict myself to PC rules I would be unable to do this kind of thing, and my world would be the poorer for it.

Glorthindel

2018-07-03, 05:38 AM

MrStabby

2018-07-03, 07:40 AM

Meh, CR is overrated. Its based on a notional player skill/optimisation level that doesn't exist in the field. It doesn't account for Magic Items, it doesn't account for rolled stats, it doesn't account for multiclassing (hell, class balance doesn't account for multiclassing, so how can CR), and it doesn't account for action economy. CR varies from a good yardstick to horribly inaccurate from level to level, and creature to creature within a level (Pixies anyone). It serves a purpose for getting a DM up and running at the start, but that is all it is good for.

We didn't need CR before 3rd ed, and other game systems get along fine without it. You don't need to assign a CR for a custom encounter, if you have been DMing for any length of time you should easily be able to tell how difficult it will be for your group. And don't think you need to know it for XP either - if you aren't using milestones anyway (where individual encounter xp is irrelevant), just give out what you normally do for an encounter, and call it good.

Oh yeah, CR is pretty much garbage for balancing encounters. I won't hate on WotC for this - it is a rough guide, a very rough guide but it doesn't and can't factor in everything (or indeed most of the important things) that determines the difficulty of a fight (for example what races/classes the PCs are or whether the conflict takes place at night or during the day).

Pex

2018-07-03, 07:47 AM

Clearly the correct balance is 1/100 classed characters in dangerous frontier areas, and 1/1000 in civilized areas. With 1/2 level 1, and 1/2 of the remainder level 2, and so on.

I can't even remember where I got this scheme from any more, but it's my basic assumption for "correct" D&D population distribution now. Even when it doesn't fit with the assumptions of an edition or campaign setting. :smallamused:

Touche, but even in 3E it was still up to the DM. Gameworld design is DM territory, but it doesn't hurt for the game to provide examples. It's rules directly affecting the PCs that need circumspection because only they have to follow them.

PhoenixPhyre

2018-07-03, 08:07 AM

I don't see how creating each NPC from scratch, going through the whole multi-step process including assigning level-ups and ASIs and rebalancing the abilities vs the party is not a significant time cost compared to taking an NPC stat block and tweaking a property or two (which usually has no affect on CR).

And even if it's not much extra work, you still have all the other downsides I mentioned. Not to mention that you're already using different build rules for monsters (unless you're totally insane), so doing it for NPCs isn't a big stretch since they play the same narrative role--as foils for the party.

Remember--5e assumes that the mechanics are for use in adjudicating actions that affect the game level, specifically interactions between the player characters and the surrounding world. They're not designed to be used to model the running of the whole world. There's no Profession (X) skills. No craft checks for NPCs (or for PCs either). If a NPC needs to have a particular ability or capability to make the story work, they just have that ability. This isn't deviating from the rules, it's obeying the rules. There's no presumption that NPCs (including monsters) and PCs work the same--any such presumption is something you're bringing in yourself.

As for commonality of magic, the DMG discusses that on pages 9 and 23-24. The default is "known but not common."

The World is Magical. Practitioners of magic are relatively few in number, but they leave evidence of their craft everywhere...

Oh yeah, CR is pretty much garbage for balancing encounters. I won't hate on WotC for this - it is a rough guide, a very rough guide but it doesn't and can't factor in everything (or indeed most of the important things) that determines the difficulty of a fight (for example what races/classes the PCs are or whether the conflict takes place at night or during the day).

That's not actually what it's for at all. CR gives a good indication of whether a particular creature (not encounter, creature) will be

a) likely to die within one round under normal conditions (defensive CR << level)
b) likely to eliminate a weak PC in one round (offensive CR >> level).

The DPR breakpoints exactly match up to the expected health of a d6 HD PC at CR = level + 1. So a oCR 2 creature is capable (in best-case scenarios where everything hits) of reducing a full-health, CON 14, level 1 wizard to zero in one round.

For balancing encounters, you use adjusted XP. And that's a pretty good guide. A strongly conservative one, by design. That's to minimize risk of unintentional TPKs.

And there are explicit instructions about balancing encounters in the DMG and in Xanathar's. Those take into consideration things like

a) do the proposed enemies primarily target the party's weak defenses?
b) terrain and other situational advantages
c) numbers (action economy balance)
etc.

MrStabby

2018-07-03, 08:59 AM

Kadesh

2018-07-03, 09:11 AM

Unoriginal

2018-07-03, 09:13 AM

PhoenixPhyre

2018-07-03, 09:13 AM

I say don't use CR to balance encounters and you respond by saying don't use CR to balance encounters? Has there been a misunderstanding?

Although you comment on XP seems odd as there is a direct relationship between CR and XP (roughly quadratic by a quick plot) so any formula you built based on XP could be converted into CR anyway.

It seemed like you were making CR out to be useless because it doesn't balance encounters. My response was that it's not supposed to balance encounters--it's useful for other reasons. It's like using a fork as a shovel. You're using it out of the design envelope, so of course it's not that useful for that.

Note the adjusted in adjusted XP. You can reach the same adjusted XP totals (within normal variation) in a whole bunch of different ways.

Consider the case of one of my parties: 3 level 2 PCs.

A medium encounter is 300-450 aXP. A hard encounter is 450-600 aXP.

Possible medium encounters:
4 CR 1/4 creatures (400 aXP)
7 CR 1/8 creatures (437 aXP)
2 CR 1/4 and 4 CR 1/8 (400 aXP)
1 CR 1 and 1 CR 1/4 (375 aXP)
etc.

And for higher levels, the possible encounters grow combinatorially. So taking the average CR doesn't do much. But accounting for
a) number/level of PCs
b) number of enemies
c) situational factors (there are instructions on how to adjust the tables for things like surprise, etc)
etc.

gives you a much better view of how much a party can take before needing a long rest.

Remember, if you're still thinking of the 3e "one or two big monsters" paradigm, you're doing 5e a disservice. APL and average CR aren't what you should look at at all. They didn't work back then, they don't work now. aXP, plus a few eyeball adjustments for the exact party, gives a much better look at things while also accounting for action advantage.

Solos without legendary actions often need to be so far above the party that they get super swingy. If they hit, they KO a PC. For me, running things entirely around that d20 roll isn't fun, since tactics don't really matter there, only luck. Solos with legendary actions, at higher levels (where there's more margin for error) can have lower relative CRs but aren't as strong as they seem to be. This is especially true if the party is at full resources.

PhoenixPhyre

2018-07-03, 09:24 AM

Can anyone walk me through the maths of the CR calcs? I'm a bit dense, and I'm wanting to throw a couple of monsters some place in a campaign I'm writing, based on an obscure early 2000's Video Game.

Fire Giant, with 5 Barbarian levels (Desert Storm Archetype). Gains Great Weapon Master instead of ASI. Replaces Ability Scores with those of a Fire Giant Dreadnought. Loses Armour and deals Bludgeoning Damage rather than Slashing with an appropriately sized Uprooted Tree. Treating Extra Attack as not stacking with MultiAttack.

Later claims a Sentient Greataxe sized to a Fire Giant (damage scale, reach etc) , which grants him the ability to cast Green Flame Blade using Con Mod up to his Reach as a Bonus Action, even when Raging.

2nd Character
Half Dragon vHuman Hexblade 17/Battle Master 3, with Polearm Master, Magic Initiate, Sweeping Strike, and 2 Epic Boons (1st Level Earth Tremor doesn't use Spell Slot, Earth Tremor doesn't use spell slot).

What CR would those characters be?

Insufficient information.

Steps to calculate CR:

Defensive CR
1) Find total health => HP
2) Adjust HP for factors from tables on page 280 and 277 of the DMG => aHP
3) Find the relevant CR in table on DMG 274 for this aHP number => base dCR
4) Find the AC for that base dCR in that same table => base AC
5) Adjust real AC for factors from tables on page 280 and 277 of the DMG => aAC
6) final dCR = base dCR + 0.5*(aAC - base AC), rounding down [technically this is by steps, but for dCR > 1, it works out the same]

Offensive CR
1) Calculate the average of the best 3 rounds of damage, assuming everything hits and limited-use abilities get used once. AoE abilities hit 2 people and they fail their saves. Take average damage. => DPR
2) Adjust DPR for factors from tables on page 280 and 277 of the DMG => aDPR
3) Find the relevant CR in table on DMG 274 for this aDPR number => base oCR
4) Find the ATK bonus (or Save if that's its dominant attack method) for that base oCR in that same table => base ATK
5) Adjust real ATK/Save for factors from tables on page 280 and 277 of the DMG => aATK
6) final oCR = base oCR + 0.5*(aATK - base ATK), rounding down [technically this is by steps, but for oCR > 1, it works out the same]

Final CR
Average final oCR and final dCR together.

Playtest
A key component--some things are weaker (or stronger) than their raw numbers suggest. See discussion on DMG 275. And bundles of stats don't define fun. Make sure it actually works as expected. A lot of the time, extra abilities get wasted. The expectation is that fights last 3-5 rounds. So that means they'll act 3-5 times on average. This does not leave lots of space for complex abilities and tactics, so making a monster dependent on them will result in frustration and waste.

Xetheral

2018-07-03, 09:27 AM

Knowing how dangerous a thing is for a group is just as important for a CaW game for exactly the reason you noted: to know how to telegraph potential danger. I don't design encounters for a specific party, but I do generally design adventuring areas for a rough difficulty. And then telegraph its danger to the party based on their relative party level.

If I didn't know the difficulty, I couldn't telegraph it, and players would wipe even more often than they already do, because they'd be acting blind without any information.

I don't agree that one needs to know CR to gauge difficulty. Just looking at the stats and eyeballing it works well enough for an initial pass for your own use. CaW expects the PCs to use off-sheet resources (e.g. allies, terrain, social engineering) to tilt combats in their favor (or make overwhelming encounters winnable), and their strategic decisions tend to control the actual difficulty a whole lot more than the enemy's abilities. All you need to telegraph is some idea of what those abilities are, so that the players can make informed strategic choices.

If you're only telegraphing CR (instead of enemy capabilities) that doesn't give the players much to work with when planning strategy. It also doesn't let the PCs make their own informed judgement about difficulty--it's simply giving them the system's (or DM's) opinion of difficulty. By contrast, when building NPCs as PCs one can let the PCs observe the enemy's use of an ability granted at a specific level of a class and let the PCs make their own inferences about the NPC's other abilities and their own judgement about how dangerous it is.

As a player, being able to make my own judgements based on IC observations is a lot more satisfying (and immersive) than knowing that the NPC could have any random set of abilities and thus needing to rely on the DM to telegraph an abstract concept like CR.

I don't see how creating each NPC from scratch, going through the whole multi-step process including assigning level-ups and ASIs and rebalancing the abilities vs the party is not a significant time cost compared to taking an NPC stat block and tweaking a property or two (which usually has no affect on CR).

And even if it's not much extra work, you still have all the other downsides I mentioned. Not to mention that you're already using different build rules for monsters (unless you're totally insane), so doing it for NPCs isn't a big stretch since they play the same narrative role--as foils for the party.

I was comparing making an NPC as a low-resolution PC (and ignoring CR) vs writing up a new statblock from scratch (and calculating CR). For the NPC you can pick build (levels, subclass(es), spells/abilities/feats), stats, skill proficiencies, and some equipment and you're done--all the other stats are derived from those choices. In contrast, for a brand-new statblock there are very few derived stats and thus a lot more choices to make individually. For example, when building an NPC as a PC, I don't need to choose the HP, because it's derived from the build. For a statblock, I need to choose the HP outright, which involves weighing mechanics (e.g. how long do I want this monster to last) and/or in-game comparisons (e.g. how many HP does it make sense for this NPC to have in comparison to other NPCs I've used previously). Then I need to repeat for AC, saves, etc. Then I still need to select spells/abilities/feats.

As for the other downsides, the only one I see as a downside is a lessened ability to rely on CR. Sure, that's a downside for DMs who want to use CR. But it doesn't matter at all to those of us who don't touch it.

Tanarii

2018-07-03, 09:40 AM

Encounter difficulty in 5e is a remarkably accurate gauge of how hard an encounter will be. It's a better basis to use for setting the mood telegraphing than "eyeballing".

Players observe abilities and make plans around them isn't "telgraphing" in the slightest. That's just players using good player skills and scouting properly. And it doesn't matter if your NPCs are PC-statted or Monster-statted for that. Observable abilities are observed regardless of underlying statting.

Also, if your hypothesis was true, players would be at a remarkable advantage vs PC-statted NPCs over regularly monsters. Just another reason not to do it.

list of reasons not to PC-class NPCs:
- faster to get stats as a DM, unless they are unique creation
- faster to run as a DM
- more accurate gauge of difficulty with less effort
- only one subsystem to think about as a DM
- designs to work with encounter paradigm instead of fight it
- allows far more NPC variety (debatable?)

Reasons to do it:
- henchmen
- DM gets to play Pcs too!
(Tongue in cheek. :smallamused: Provide your list here. But so far I'm not covinced of any other reasons.)

PhoenixPhyre

2018-07-03, 09:42 AM

I was comparing making an NPC as a low-resolution PC (and ignoring CR) vs writing up a new statblock from scratch (and calculating CR). For the NPC you can pick build (levels, subclass(es), spells/abilities/feats), stats, skill proficiencies, and some equipment and you're done--all the other stats are derived from those choices. In contrast, for a brand-new statblock there are very few derived stats and thus a lot more choices to make individually. For example, when building an NPC as a PC, I don't need to choose the HP, because it's derived from the build. For a statblock, I need to choose the HP outright, which involves weighing mechanics (e.g. how long do I want this monster to last) and/or in-game comparisons (e.g. how many HP does it make sense for this NPC to have in comparison to other NPCs I've used previously). Then I need to repeat for AC, saves, etc. Then I still need to select spells/abilities/feats.

As for the other downsides, the only one I see as a downside is a lessened ability to rely on CR. Sure, that's a downside for DMs who want to use CR. But it doesn't matter at all to those of us who don't touch it.

But the sheer number of calculations are much larger. You have to do level-by-level calculations and account for interactions (especially with races and multi-classing).

I calculated that a level 1 PC makes between 8 and 14 long-term mechanical decisions, and that's under-counting tremendously. And then 2-4 more per level after that.

For a "stat block" creature I can do the following:
1) choose an oCR. This gives me a DPR range and an attack bonus.
2) choose an dCR. This gives me a HP range and the AC.
3) Pick the rest arbitrarily to match the theme of the NPC. Because most things don't really matter as long as they're in a standard range.

Or even easier, pick a base stat block and add/change a few things (damage types, vulnerabilities, spells, etc). Takes 10s (can be done on the fly during play). That's an option you just don't have when you build everything from scratch as a PC. You don't have a fall-back, everything has to be from scratch.

MrStabby

2018-07-03, 10:07 AM

It seemed like you were making CR out to be useless because it doesn't balance encounters. My response was that it's not supposed to balance encounters--it's useful for other reasons. It's like using a fork as a shovel. You're using it out of the design envelope, so of course it's not that useful for that.

But I have never used CR for balancing encounters, never recommended CR for balancing encounters never have I suggested that anyone else suggested that CR be used to balance encounters... but yeah I said it was rubbish and implied it shouldn't be used. Not sure how you got "You're using it out of the design envelope" from that.

Now I also quite happy to say that other things shouldn't be used to balance encounters - XP and Adjusted XP (or adjusted for number of monsters anyway - other adjustments are better). Both are based on CR and both just give a false sense of control.

PhoenixPhyre

2018-07-03, 10:11 AM

But I have never used CR for balancing encounters, never recommended CR for balancing encounters never have I suggested that anyone else suggested that CR be used to balance encounters... but yeah I said it was rubbish and implied it shouldn't be used. Not sure how you got "You're using it out of the design envelope" from that.

Now I also quite happy to say that other things shouldn't be used to balance encounters - XP and Adjusted XP (or adjusted for number of monsters anyway - other adjustments are better). Both are based on CR and both just give a false sense of control.

But bashing something as being bad at X presumes that you want it to be used for X. Bashing a fork for being a bad shovel is just :sideways_owl: territory.

In my experience, aXP works very well to predict the resources used in a straight fight. It's definitely a conservative measure, providing a ceiling rather than a floor, but that's by design (and good design, in my opinion).

Unoriginal

2018-07-03, 10:36 AM

Now I also quite happy to say that other things shouldn't be used to balance encounters - XP and Adjusted XP (or adjusted for number of monsters anyway - other adjustments are better). Both are based on CR and both just give a false sense of control.

A few days ago I DMed a couple of fights using the DMG's and XGtE's guidelines and calculations to test someone's houserule, the results of the 1 vs 1 fight landed squarely within what the books said it'd end up as, even with the houserule taken into account.

Same can be said of any other fights I've run or seen, as far as I recall.

It's not a "false sense of control", it's a system working as indicated

That being said, I'd be curious how you balance an encounter for, say, a Fighter and an Barbarian, both lvl 7, vs NPCs built like you do.

MrStabby

2018-07-03, 10:37 AM

But bashing something as being bad at X presumes that you want it to be used for X. Bashing a fork for being a bad shovel is just :sideways_owl: territory.

In my experience, aXP works very well to predict the resources used in a straight fight. It's definitely a conservative measure, providing a ceiling rather than a floor, but that's by design (and good design, in my opinion).

Actually I would say that "bashing something as being bad at X" presumes that your interlocutor is wanting it to be used for X. If I am trying to dissuade someone from doing something it is more likely to be because I believe they want to do it than because I do.

That aside, aXP is a little better but not much when it comes to balancing encounters. As much impact comes from class composition, environment, spells known (and resources left) and so on as it does from the enemies themselves. It is the matching of the abilities of the party against the strengths of the enemies that matters - any system that only looks at one half of the relationship is always going to be doomed. To be fair it does look at the PC side, but "number of PCs" and "Level" is too much of an oversimplification. As an extreme example consider a party of four clerics against a horde of low level undead - four sets of destroy undead (or 8 if level 6) will make it a trivial encounter; for a party of 4 warlocks that same encounter is likely to leave them devoid of resources and in need of a rest, if they survive it at all.

Unoriginal

2018-07-03, 10:39 AM

But bashing something as being bad at X presumes that you want it to be used for X. Bashing a fork for being a bad shovel is just :sideways_owl: territory.

Well, it presumes you think other people want it to be used like that and that you disagree.

PhoenixPhyre

2018-07-03, 11:07 AM

Actually I would say that "bashing something as being bad at X" presumes that your interlocutor is wanting it to be used for X. If I am trying to dissuade someone from doing something it is more likely to be because I believe they want to do it than because I do.

That aside, aXP is a little better but not much when it comes to balancing encounters. As much impact comes from class composition, environment, spells known (and resources left) and so on as it does from the enemies themselves. It is the matching of the abilities of the party against the strengths of the enemies that matters - any system that only looks at one half of the relationship is always going to be doomed. To be fair it does look at the PC side, but "number of PCs" and "Level" is too much of an oversimplification. As an extreme example consider a party of four clerics against a horde of low level undead - four sets of destroy undead (or 8 if level 6) will make it a trivial encounter; for a party of 4 warlocks that same encounter is likely to leave them devoid of resources and in need of a rest, if they survive it at all.

OK, let's pose the question back at you. What do you use to balance encounters? By your standards, looking at the creatures and their statistics is tantamount to using CR, since CR is a straight-forward calculation from those stats, just like aXP is a straight-forward calculation from CR.

Do you tell a new DM to "just eyeball it"? Eyeball what exactly? Simulate the whole encounter? That's extremely labor intensive and error-prone (tactics matter). So what do you do?

MrStabby

2018-07-03, 11:50 AM

OK, let's pose the question back at you. What do you use to balance encounters? By your standards, looking at the creatures and their statistics is tantamount to using CR, since CR is a straight-forward calculation from those stats, just like aXP is a straight-forward calculation from CR.

Do you tell a new DM to "just eyeball it"? Eyeball what exactly? Simulate the whole encounter? That's extremely labor intensive and error-prone (tactics matter). So what do you do?

For a New DM you ask? Well start with a published module. Not all are perfect but this is a pretty good place to get the hang of how the system works and as some encounters are relatively easy/difficult have a think about what makes them so. Get a sense of what has an impact. As you say, tactics matter (i.e. CR/XP is a terrible way to gauge difficulty). Once you have a bundle of other peoples encounters under your belt then design your own.

The design also depends on game type and style. If you are playing a brutal game and retreat is a valid tactical option in a sandbox type world then you do have an easier time - if your encounter is too hard you have a whole team of people onside to turn it into an evasion encounter rather than a combat encounter. If it is too easy then little is lost - learn from it and move on. If you are on Heroic/Easy setting for your campaign then I would advise them to still eyeball it but tone it down a bit. Ballpark estimate for raw armour/HP is to take 5 rounds of hitting away to kill everything off - at least at low levels (up to about 7) (this also assumes roughly equal numbers on each side - being able to kill of enemies to lower damage is advantageous, but this is a rough process). Then the PCs will use resources and the NPCs will do the same to speed up and slow down the encounter. Offence from the NPCs should also take about 5 rounds to kill the players (discounting tactics, resource usage etc.) so the encounter starts of balanced but the PCs should have more resources/tools to use to swing the fight - this is what takes it from deadly to moderate difficulty. Tweak as needed to manage difficulty/PC magic items/Tactics etc.. After 6 or 7 sessions like this a DM should have an even better understanding of what is going on and be able to quickly eyeball encounters.

If you don't like to tailor encounters to a party then make up an imaginary party of representative classes to test it against.

Tanarii

2018-07-03, 12:00 PM

Yeah, I'll take the system method over anyone's claims that they can "eyeball it" better any day of the week. :smallyuk:

Unoriginal

2018-07-03, 12:05 PM

PhoenixPhyre

2018-07-03, 12:08 PM

For a New DM you ask? Well start with a published module. Not all are perfect but this is a pretty good place to get the hang of how the system works and as some encounters are relatively easy/difficult have a think about what makes them so. Get a sense of what has an impact. As you say, tactics matter (i.e. CR/XP is a terrible way to gauge difficulty). Once you have a bundle of other peoples encounters under your belt then design your own.

The design also depends on game type and style. If you are playing a brutal game and retreat is a valid tactical option in a sandbox type world then you do have an easier time - if your encounter is too hard you have a whole team of people onside to turn it into an evasion encounter rather than a combat encounter. If it is too easy then little is lost - learn from it and move on. If you are on Heroic/Easy setting for your campaign then I would advise them to still eyeball it but tone it down a bit. Ballpark estimate for raw armour/HP is to take 5 rounds of hitting away to kill everything off - at least at low levels (up to about 7) (this also assumes roughly equal numbers on each side - being able to kill of enemies to lower damage is advantageous, but this is a rough process). Then the PCs will use resources and the NPCs will do the same to speed up and slow down the encounter. Offence from the NPCs should also take about 5 rounds to kill the players (discounting tactics, resource usage etc.) so the encounter starts of balanced but the PCs should have more resources/tools to use to swing the fight - this is what takes it from deadly to moderate difficulty. Tweak as needed to manage difficulty/PC magic items/Tactics etc.. After 6 or 7 sessions like this a DM should have an even better understanding of what is going on and be able to quickly eyeball encounters.

If you don't like to tailor encounters to a party then make up an imaginary party of representative classes to test it against.

You've basically come up with a cruder, rougher version of CR, because that's exactly what it measures. Oh, and your system requires long calibration (6-7 sessions is basically T1 or so) when the PCs are the most fragile and then has to be regularly re-calibrated, while the aXP works 99% of the time out of the box and fails on the safe side when it fails. So no. You've reinvented the wheel, just rectangular.

The aXP part already takes into account party composition and available tactics/weak saves, etc.

MrStabby

2018-07-03, 12:15 PM

So you're saying "you have to calculate or eyeball the HPs, AC, accuracy and damage output of all the actors in the encounter, several of them you had to create from scratch using the PC rules, then you have to take into account environment, tactics and matchup issues between the foes, possibly also run the encounter with a test-PC group to see if it works, so that each group can kill the other in a given number of rounds"

And somehow, " use XPs and CRs (which are based on HPs, AC, accuracy and damage output) then adjust according to factors like environment, foe matchups and the like" is the option that's less accurate/too complicated/not sufficient?

Ok.

I find it particularly fun that when prompted, your first advice for new DMs is "use published modules", whose NPCs are definitively not built like PCs.

Maybe your experience is that CR is better than your attempts to eyeball encounters. Mine is the opposite. When I relied on judgement rather than an oversimplified formula my encounters got better, not worse. When you did the same your experience was reversed - that's fine. If a formula is better than your judgement then use the formula; use whatever works best for you.

Also I am not understanding why you bring up NPCs not being built like PCs? That seems tangential at best and I wan't recommending building NPCs like PCs? Can you clarify what this is in response to?

Unoriginal

2018-07-03, 12:21 PM

Also I am not understanding why you bring up NPCs not being built like PCs? That seems tangential at best and I wan't recommending building NPCs like PCs? Can you clarify what this is in response to?

I apologize, then. You agreed with Glorthindel about CR, and their dissatisfaction about CR was part of their argument to use PC rules for NPCs, so I assumed you shared this view as well. It was wrong of me to do so.

Pex

2018-07-03, 12:25 PM

Yeah, I'll take the system method over anyone's claims that they can "eyeball it" better any day of the week. :smallyuk:

If you apply this to skill use . . .

Unoriginal

2018-07-03, 12:32 PM

If you apply this to skill use . . .

Very easy DC 5
Easy DC 10
Medium DC 15
Hard DC 20
Very hard DC 25
Nearly impossible DC 30

What are you supposed to eyeball? The system for ability check use is clear.

Malifice

2018-07-03, 12:36 PM

NPCs dont have levels. Only PCs do.

MrStabby

2018-07-03, 12:37 PM

I apologize, then. You agreed with Glorthindel about CR, and their dissatisfaction about CR was part of their argument to use PC rules for NPCs, so I assumed you shared this view as well. It was wrong of me to do so.

Hey, no problem. Things get mixed up on threads. Just some confusion.

Tanarii

2018-07-03, 01:04 PM

If you apply this to skill use . . .
To misused a quote, a foolish consistency something something.

Edit: just in case, imagine me looking haughty and waving my hands dismissively, am elderly matron caught out by a good point but refuses to admit it. :smallamused:

Unoriginal

2018-07-03, 02:23 PM

To misused a quote, a foolish consistency something something.

Edit: just in case, imagine me looking haughty and waving my hands dismissively, am elderly matron caught out by a good point but refuses to admit it. :smallamused:

It's not a good point, though.

Tanarii

2018-07-03, 02:37 PM

It's not a good point, though.
Aye. The two are comparable in that they depend somewhat on the DM's good judgement within the broad guidelines provided.

But Pex's sarcastic comment does help me understand that to some people, the broad guidelines are going to be too broad to where they consider them to be useful. And to others, they'll consider their own way of doing things superior to the guidelines provided, and that it won't cause them any trouble to just ignore them.

It's people that consider themselves to be experiencing a problem when they insist on ignoring the guidelines provided that it's worth pointing that out to. Not people that say they personally have a problem when using them, and prefer an alternative way of doing things. Just because I find it more problematic to ignore them than use them doesn't mean everyone will.

Or possibly I've just had too much coffee and I'm rambling now.

Unoriginal

2018-07-03, 03:08 PM

Aye. The two are comparable in that they depend somewhat on the DM's good judgement within the broad guidelines provided.

Not quite.

Establishing a DC for a task based on its difficulty is following the guidelines.

Eyeballing an encounter's difficulty without doing the calculation is disregarding the guidelines

Nothing wrong with not following the guidelines, but the two cases are different.

It's people that consider themselves to be experiencing a problem when they insist on ignoring the guidelines provided that it's worth pointing that out to. Not people that say they personally have a problem when using them, and prefer an alternative way of doing things. Just because I find it more problematic to ignore them than use them doesn't mean everyone will.

Fair.

Pex

2018-07-03, 06:37 PM

Very easy DC 5
Easy DC 10
Medium DC 15
Hard DC 20
Very hard DC 25
Nearly impossible DC 30

What are you supposed to eyeball? The system for ability check use is clear.

Way to miss the point, but to discuss further is off topic.

Xetheral

2018-07-03, 10:42 PM

Encounter difficulty in 5e is a remarkably accurate gauge of how hard an encounter will be. It's a better basis to use for setting the mood telegraphing than "eyeballing".

In a Combat-as-Sport game, where mechanical stats are the primary component of difficulty, sure, maybe CR and DMG encounter difficulty are more accurate than eyeballing (I suspect an experienced DM's intuition can be similarly accurate, but I have no evidence to back that up other than my own experience, so I'll concede the point). But in a Combat-as-War game, encounter difficulty is set more by in-game-events. The PCs might manage to get the city guard as allies, or distract the lone guard and butcher the other enemies in their sleep, or make it politically too costly for the NPCs to attack, or engineer distrust among the enemy group, or lure the enemies into an avalanche or unstable structure, or call in a favor with a dragon, or countless other methods of defeating the opposition where stats don't end up mattering, frequently making CR and DMG encounter difficulty irrelevant.

Players observe abilities and make plans around them isn't "telgraphing" in the slightest. That's just players using good player skills and scouting properly.

I define telegraphing as giving the players the opportunity to learn about potential opponents far enough in advance for them to make informed decision regarding whether and how to engage or avoid the opposition. You may define it differently.

And it doesn't matter if your NPCs are PC-statted or Monster-statted for that. Observable abilities are observed regardless of underlying statting.

I disagree. Oberserving a single ability of a hodge-podge statblock only gives information about that ability. Observing a single ability of a PC-built NPC provides a list of other lower-level abilities, and a minimum floor for HP and proficiency bonus.

Also, if your hypothesis was true, players would be at a remarkable advantage vs PC-statted NPCs over regularly monsters. Just another reason not to do it.

That's what knowledge skills are useful for.

Reasons to do it:
- henchmen
- DM gets to play Pcs too!
(Tongue in cheek. :smallamused: Provide your list here. But so far I'm not covinced of any other reasons.)

You may not value any of these, but here are some things from my list:

Consistent methodology. Similar things (in the game world) are modelled similarly. Avoids modeling aritifacts. Two characters with identical life experiences and decisions have the same abilities regardless of whether they are a PC or NPC. Appearance of fairness. Avoids NPCs being able to have combinations of abilities forbidden to PCs. Makes PC observations more informative (discussed above). Increased immersion from a more-consistant game world.

For a "stat block" creature I can do the following:
1) choose an oCR. This gives me a DPR range and an attack bonus.
2) choose an dCR. This gives me a HP range and the AC.
3) Pick the rest arbitrarily to match the theme of the NPC. Because most things don't really matter as long as they're in a standard range.

This may largely be where we differ. When making a custom statblock I would go through each stat and determine what is reasonable based on the NPC's role in the game world, cross-referencing to other NPCs to ensure consistency. Then, if I we're using CR, I'd calcuate it (and then adjust the stats, like proficiency bonus, that are based on CR). Your approach would certainly save time, but I have no interest in starting with the desired CR because I build encounters based on the game world rather than based on the mechanics.

JoeJ

2018-07-04, 01:08 AM

Oberserving a single ability of a hodge-podge statblock only gives information about that ability. Observing a single ability of a PC-built NPC provides a list of other lower-level abilities, and a minimum floor for HP and proficiency bonus.

To the extent this is true, I count it as a very good reason not to build NPCs that way. How does it make sense that observing somebody picking a lock should tell you anything at all about how skillfully they wield a battleaxe, or vice versa?

JackPhoenix

2018-07-04, 07:59 AM

I disagree. Oberserving a single ability of a hodge-podge statblock only gives information about that ability. Observing a single ability of a PC-built NPC provides a list of other lower-level abilities, and a minimum floor for HP and proficiency bonus.

And that happens how? The NPC thinks "Oh, the PC's are looking at me, I have to use my most powerful ability right now, so they can guess what class and level I am? Otherwise...
"Oh, look, the evil wizard used Fireball, we can take him... WTF? Why are we all dead? Where did that Meteor Swarm came from?"

Not to mention that it's not very informative, in many cases. You see someone make 3 attacks in a turn... does that mean he's level 11 fighter, level 5 Berserker barbarian, level 20 Bladesinger under Haste, or level 3 Hunter ranger with Hordebreaker who got lucky with GWM bonus attack?

Avoids modeling aritifacts. Two characters with identical life experiences and decisions have the same abilities regardless of whether they are a PC or NPC.

That happens with the normal method too. It's just that vast majority of NPCs *don't* have identical life experiences and decisions as PCs, and thus they use different mechanics. By "avoiding modeling artifacts", you create different sort of modeling artifacts.

Increased immersion from a more-consistant game world.

Technically, world where everyone follows one of 12 possible life paths is more consistent. I wouldn't, however, find it more immersive. I'm reminded of joke from Fallout 2, based on the limited amount of character models... "It looks like there's only about 20 different people in the world".

This may largely be where we differ. When making a custom statblock I would go through each stat and determine what is reasonable based on the NPC's role in the game world, cross-referencing to other NPCs to ensure consistency. Then, if I we're using CR, I'd calcuate it (and then adjust the stats, like proficiency bonus, that are based on CR). Your approach would certainly save time, but I have no interest in starting with the desired CR because I build encounters based on the game world rather than based on the mechanics.

You still have to decide what level you want your NPC be. That's not very different from deciding what CR do you want the opponent be.

PhoenixPhyre

2018-07-04, 08:06 AM

And that happens how? The NPC thinks "Oh, the PC's are looking at me, I have to use my most powerful ability right now, so they can guess what class and level I am? Otherwise...
"Oh, look, the evil wizard used Fireball, we can take him... WTF? Why are we all dead? Where did that Meteor Swarm came from?"

Not to mention that it's not very informative, in many cases. You see someone make 3 attacks in a turn... does that mean he's level 11 fighter, level 5 Berserker barbarian, level 20 Bladesinger under Haste, or level 3 Hunter ranger with Hordebreaker who got lucky with GWM bonus attack?

That happens with the normal method too. It's just that vast majority of NPCs *don't* have identical life experiences and decisions as PCs, and thus they use different mechanics. By "avoiding modeling artifacts", you create different sort of modeling artifacts.

Technically, world where everyone follows one of 12 possible life paths is more consistent. I wouldn't, however, find it more immersive. I'm reminded of joke from Fallout 2, based on the limited amount of character models... "It looks like there's only about 20 different people in the world".

You still have to decide what level you want your NPC be. That's not very different from deciding what CR do you want the opponent be.

I second all of this. In addition, if you're using any non-player-race NPCs, you're already using "different rules" for those. So your consistency is gone already. Because none of those are built the same way PCs are (and you wouldn't expect them to--how does a manticore use weapon proficiencies that come with classes?).

It's much better to have consistency by game role for game statistics. Remember, the game statistics don't define everything the creature is capable of. They only provide a mechanical representation of those things necessary to face them in combat.

5e's rules do not make any attempt to simulate a world. They exist for the sole purpose of adjudicating actions involving PCs in a fun, playable way. That's it. What happens off-screen is not bound by those rules at all (which is good, because any playable set of rules is completely incapable of providing a satisfying world simulation. The two goals are incompatible at the core.)

Tanarii

2018-07-04, 09:23 AM

5e's rules do not make any attempt to simulate a world. They exist for the sole purpose of adjudicating actions involving PCs in a fun, playable way. That's it. What happens off-screen is not bound by those rules at all (which is good, because any playable set of rules is completely incapable of providing a satisfying world simulation. The two goals are incompatible at the core.)
You're not going to get very far with that, although I completely agree that's the design intent for 5e. But Xetheral has made it clear he believes the rules are meant to simulate the in-game universe in several recent threads.

Xetheral

2018-07-04, 09:49 AM

To the extent this is true, I count it as a very good reason not to build NPCs that way. How does it make sense that observing somebody picking a lock should tell you anything at all about how skillfully they wield a battleaxe, or vice versa?

It doesn't make any sense. But observing tool proficiency use doesn't provide that kind of information anyway, so it's a moot point.

And that happens how? The NPC thinks "Oh, the PC's are looking at me, I have to use my most powerful ability right now, so they can guess what class and level I am? Otherwise...
"Oh, look, the evil wizard used Fireball, we can take him... WTF? Why are we all dead? Where did that Meteor Swarm came from?"

It happens whenever the PCs observe the NPCs in combat, or, when doing legwork, when they track down and talk to others who have observed the NPCs in combat. In the case of spellcaster, it also happens when observing (or talking to anyone who has observed) utility spell use.

And yes, observations provide a minimum only. That's still more than observing an NPC statblock where observations permit even fewer inferences.

Not to mention that it's not very informative, in many cases. You see someone make 3 attacks in a turn... does that mean he's level 11 fighter, level 5 Berserker barbarian, level 20 Bladesinger under Haste, or level 3 Hunter ranger with Hordebreaker who got lucky with GWM bonus attack?

Still more information than observations of an NPC statblock.

That happens with the normal method too. It's just that vast majority of NPCs *don't* have identical life experiences and decisions as PCs, and thus they use different mechanics. By "avoiding modeling artifacts", you create different sort of modeling artifacts.

The vast majority of NPCs never need to be modelled mechanically at all, so no additional artifacts are created. And I avoid artifacts such as a thieves-guild-trained PC somehow having entirely different abilities than their fellow students.

Technically, world where everyone follows one of 12 possible life paths is more consistent. I wouldn't, however, find it more immersive. I'm reminded of joke from Fallout 2, based on the limited amount of character models... "It looks like there's only about 20 different people in the world".

First, I allow (and promote) multiclassing. So your number of 12 is off by quite a few orders of magnitude. Second, there isn't a one-to-one correspondence between "life paths" and mechanical builds, so the comparison is flawed in the first place. I want similar life experiences to be modelled similarly--that doesn't prevent dissimilar life experiences from also being modelled similarly. (It is a model, after all. I expect to lose complexity.)

You still have to decide what level you want your NPC be. That's not very different from deciding what CR do you want the opponent be.

This is a good point and one I hadn't considered. I'll think about it some more.

I second all of this. In addition, if you're using any non-player-race NPCs, you're already using "different rules" for those. So your consistency is gone already. Because none of those are built the same way PCs are (and you wouldn't expect them to--how does a manticore use weapon proficiencies that come with classes?).

I'm looking for consistency, not balance. Both PC and NPC races get racial abilities from their species and class abilities from their class. I don't mind that the NPC racial abilities aren't in line with PC racial abilities.

It's much better to have consistency by game role for game statistics. Remember, the game statistics don't define everything the creature is capable of. They only provide a mechanical representation of those things necessary to face them in combat.

You and I value different things when it comes to mechanics. I agree that consistency by game role is "much better" for your table. I disagree that it is better at mine.

5e's rules do not make any attempt to simulate a world. They exist for the sole purpose of adjudicating actions involving PCs in a fun, playable way. That's it. What happens off-screen is not bound by those rules at all (which is good, because any playable set of rules is completely incapable of providing a satisfying world simulation. The two goals are incompatible at the core.)

I'm not trying to simulate a world with the rules. I'm trying to consistently model how PCs and NPCs fight: both what actions they are capable of and how those actions will be resolved. That effort also provides consistency in modeling utility spells, which is an extra benefit. Combat (and utility spell) modeling is a far cry from world simulation.

Edit:

You're not going to get very far with that, although I completely agree that's the design intent for 5e. But Xetheral has made it clear he believes the rules are meant to simulate the in-game universe in several recent threads.

To clarify, I view the game at my table as an imagination-based (rather than mechanical) simulation of the PCs in a fantasy world. I find that simulation more satisfying when the mechanical model used to resolve certain parts of the simulation (e.g. combat) gives results that are consistent with the simulation as a whole.

I do not expect or believe that 5e was specifically designed to cater to my preferences. I do, however, think 5e deliberately supports my preferences as part of it's big-tent approach, as evidenced by the inclusion of the rule for creating NPCs as PCs.

MrStabby

2018-07-04, 11:04 AM

A lot of this isn't mutually exclusive, some of it is.

I do agree with some of both sides on this.

Building NPCs as PCs is usually not a good idea. PCs are complex and are designed to function across an adventuring day - in and out of combat abilities for example. They are also designed to work in parties of about 3 to 5 people. If you build NPCs as PCs you tend to get more offensively powerful enemies and defensively weaker ones. This also ties your hands somewhat - if an encounter would benefit from an enemy that can twin a spell from the druid list then it is just easier to let them do that than work out a sorcerer/druid multiclass and how low its constitution must be to have the HP appropriate for that encounter what with all the extra class levels you needed to give it.

On the other hand it is really useful to have world consistency so the players can infer things about the world from what they see. I like that if I see a caster lay down a wall of force I can deduce they probably have a good save against illusions but a much less good dexterity save. I like that if I see a big warrior with a prominent holy symbol on their shield with multiple attacks that casts spells like wrathful smite I might guess they are helping the saves of those around them.

The point is that you can do both. Just because you don't follow the rules for making a PC when designing an NPC doesn't stop you from giving the NPC abilities that would be thematic for a class. You can still build consistency into the world - sure it is a bit different from PC consistency but that isn't the end of the world.

Likewise with breaking the "rules" - you can still do that and signal to the PCs that things are different - maybe the druid twinning spells is an Ettin and one effect comes from each head - there is something there immediately suggesting why this guy is different (in in your world that could be something common to all Ettin spellcasters). Tales, lore and other indicators can all play a part in helping some of these things be anticipated by smart players, history and other knowledge checks can help fill in blanks.

Ultimately it all still comes down to what makes the most fun game, which is a table specific thing. If a table doesn't like something adapt to that table. It isn't a choice between two extremes but a bit of a spectrum and you chose where on that spectrum you want your game to be. Does tying your hands bring more pleasure to the group with enforced consistency than you lose through optimising elements of your campaign world for the fun they bring to the combat?

Requilac

2018-07-04, 11:23 AM

To bring this back on topic...

The main difference between a PC and an NPC/monster that you need to keep in mind is that PC's are balanced around the idea that they will be expending resources over multiple fights, while an NPC/monster is unlikely to survive its first fight of the day. If the DM has to reference a NPC's/monsters stat block for more than a single less than one hour setting, it is a surprise. PCs are not supposed to fight just one enemy a day, a mistake that I made as a DM not too long ago and it upset the balance of the game greatly.* Meanwhile a player will be referencing their PC statblock for combat after combat after combat. This creates a situation where a feature given to a monster would be incredibly overpowered if given to a player, or vice versa.

Take the Doppelganger's Read Thoughts as an example. Because it belongs to a monster its really not that great. And as a DM my firts reaction is that it is cool, but altogether unlikely to be that useful. It would typically only use this feature out of combat, and a lot of times I would just hand waive how it learned that and not detail whose mind it read. I would just say that it knew the location of the diseased king's hiding place (or whatever) because of this feature, which I could also explain a bunch of different ways. Its basically just a flavor thing for the Doppelganger. Now imagine me giving this to a player. Getting to read anything's mind without a chance of failure or expending resources, and nearly constant advantage on nearly all charisma ability checks? Way overpowered.

And also, keep in mind that hit dice is based on a monster's size, not its class. Check page 276 of the DMG. And what you also need to remember is that the monster's con modifier is added to each hit die roll, so sturdier monsters tend to actually have less hit dice than you would expect. If two monsters have the same hit points, but different constitution modifiers, the tougher one with a higher modifier will actually have less hit dice.

As for the level of magic in the world, that is dependent on the Campaign Setting. Even in existing settings there will be a drastic difference in levels of magic. This varies wildly and there is no solid answer. The Forgotten Realms is fairly high magic and weaker magic items are fairly easy to buy, but even in FR there is not much consistency.

*Sjappo, if you are reading this I just wanted to say I should have taken your advice much earlier.

JackPhoenix

2018-07-04, 11:24 AM

It happens whenever the PCs observe the NPCs in combat, or, when doing legwork, when they track down and talk to others who have observed the NPCs in combat. In the case of spellcaster, it also happens when observing (or talking to anyone who has observed) utility spell use.

And I suppose you tell them the exact abilities they use? Because "We've send a squad of guys against the renegade wizard, and he burned them all with blasts of fire" doesn't tell you anything about his real capabilities. No matter if he's build as PC or using NPC statblock. And I wouldn't expect players need to "research" every opponent they come across before they face him in battle.

Still more information than observations of an NPC statblock.

Not really. "He's a guy who can attack pretty fast" is exactly the same information you get from NPC statblock. In either case, it doesn't actually tell you anything about how dangerous that guy is, or what else can he do.

The vast majority of NPCs never need to be modelled mechanically at all, so no additional artifacts are created. And I avoid artifacts such as a thieves-guild-trained PC somehow having entirely different abilities than their fellow students.

How so? You've never had the PCs pick a fight with random NPCs before? If the party barbarian randomly decide to start a bar brawl, I simply decide that most people there are commoners, there may be some better brawlers who are best to be modelled by bandits or guards, and there's the local tough and bully, who's good enough to be thug. About 5 seconds of work. If I really want to, I may give them more HP or modify their stats on spot, which takes about 5 seconds more. You'll have to stat them up like PCs, with all that involve.

And as for "entirely different abilities", unless every student is clone of each other, that will happen too. Need a "rogue"? Spy statblock got my back. Need better rogue? Assassin. Or worse rogue? I can simply add 1d6 sneak attack to a bandit statblock. Amount of work and time involved? Again, about 5 seconds. And the result is similar enough for any actuall in-game needs.

First, I allow (and promote) multiclassing. So your number of 12 is off by quite a few orders of magnitude. Second, there isn't a one-to-one correspondence between "life paths" and mechanical builds, so the comparison is flawed in the first place. I want similar life experiences to be modelled similarly--that doesn't prevent dissimilar life experiences from also being modelled similarly. (It is a model, after all. I expect to lose complexity.)

And the NPCs are similar. Mage is similar to 9th level wizard, but without all that unnecessary complexity creating a level 9 wizard requires. Same with Assassin and rogue, or Veteran and experienced fighter.

I'm looking for consistency, not balance. Both PC and NPC races get racial abilities from their species and class abilities from their class. I don't mind that the NPC racial abilities aren't in line with PC racial abilities.

So why do you have problem with "NPC class" abilities being different from PC class abilities? Considering you care about this while "stating NPCs as PCs", I assume you're familiar with 3.x NPC classes. They are similar to PC classes, but no sane player would choose to play one. Using a Thug statblock instead of a level 5 fighter isn't different from using a level 5 Warrior instead of level 5 fighter. "Thug" is simply a member of NPC class with d8 hit die, pack tactic, extra attack (yeah, screw multiattack limitations, I agree with that), and no other class features.

Pex

2018-07-04, 11:36 AM

PhoenixPhyre

2018-07-04, 11:39 AM

yeah, screw multiattack limitations, I agree with that

Multiattack and Extra Attack are not the same feature. Nor are they supposed to be.

Extra Attack is a specific exception to the "when you take the attack action you make one attack" rule. It has specific considerations how it stacks (it doesn't, unless you're a fighter and then only with the fighter features) and what you can do (only make more weapon attacks or things that directly substitute for weapon attacks during the Attack action).

Multiattack is a specific action. It's not the attack action at all. It completely ignores the usual Attack action limits and requirements, substituting a whole different set of capabilities and limitations. Among those:

a) you can do things other than make another weapon attack, but only what's specified. Some can use other actions such as a dragon's fear aura ability. Others can make much more than one attack, regardless of "level."
b) you can't do things that trigger off of the Attack action (including grappling or shoving). Creatures that should have those capabilities get them built into the attacks themselves.
c) it simplifies a DM's overhead. On each turn, an enemy can do one and only one of the things listed under its "Action" block unless it specifically says otherwise. You don't have to worry about actions and bonus actions and conditionals unless you really want to. Pick one entry, resolve. Move on to the next creature. If you build everything like a PC, you lose this advantage and you dramatically increase the overhead of the DM and thus the length of the DM's turn. Or you force DMs to use fewer enemies, which has all sorts of screwy results.
D) it doesn't remove the capability of enemies to take the regular attack action and use any weapon they possess or grapple or shove. But it does mean that they can't do that as part of a multi-attack routine.

MrStabby

2018-07-04, 11:49 AM

It's nitpicking, but you have to be careful when you're too obvious giving an NPC an ability PCs normally don't have such as twinning druid spells. Players will want to see evidence that NPC is a druid/sorcerer multiclass. Minutiae of ability scores and skill proficiency won't matter because players don't experience those, but a druid twinning Heat Metal against the PC fighter's and paladin's armor will be called out. It's not difficult for the DM to do, just show during the combat the NPC casting sorcerer and druid spells. It's not logical, but not having that proof is more glaring than if a troll is vulnerable to cold and electricity instead of fire and acid. The DM wasn't wrong in having the NPC twin druid spells, but he can lose player trust if he keeps breaking the rules so to speak.

This was kind of my point about making such a caster an Ettin - you are getting the effect you wan't but you are changing the cause from "being a sorcerer" to "having two heads". As long as your campaign isn't going to include true polymorph you are probably OK with this. There is a consistent explanation: it isn't about the players not spotting it, but about it not sticking in their heads as being "unreasonable" - in the sense of inexplicable.

In my campaign I have a two headed dragon in my back-pocket waiting to bring out. I am hoping that the extra casting ability from the second head will not be deemed to be unreasonable.

PhoenixPhyre

2018-07-04, 11:55 AM

See, I make it very clear upfront that player characters and NPCs are different. There aren't "Druids" in the class-feature sense. There are people who use druidic magic, whether those are called shamans, spiritualists, witches, or whatever. Most of them can't shapeshift, some of them act as somewhat of a hybrid of classes, others are completely different. I don't pick from the "druid" list--that's not an in-universe thing. It's a game thing, for convenience and balance.

I've had dragons that eat souls, I have one NPC that can cast any 6th-level or lower spell as a wizard spell without a spellbook, giants who wield rune magic, etc. Restricting oneself to the 12 classes and the printed abilities and skills makes a low-resolution world, lacking in wonder and discovery.

As for it being supported--it is. But grudgingly. The text of the DMG makes it clear that you can stat NPCs as PCs, but if there's a simpler way it's probably better not to.

JNAProductions

2018-07-04, 12:05 PM

I can see statting a few important NPCs as PCs. I can also see not doing that, if you want them to have different abilities from players, or doing it MOSTLY, but adding or subtracting a few things here or there.

But it's 100% not needed. NPCs are built differently from PCs, and to me, that's a good thing. Gets rid of nonsense like a giant gelatinous puddle of guck having a BAB and saves in the stratosphere, simply because it's got a lot of HD.

PhoenixPhyre

2018-07-04, 12:11 PM

But it's 100% not needed. NPCs are built differently from PCs, and to me, that's a good thing. Gets rid of nonsense like a giant gelatinous puddle of guck having a BAB and saves in the stratosphere, simply because it's got a lot of HD.

Or your top blacksmiths being able to fight off an army of guards, because the craft checks needed to just do his job are that high.

Tying combat and non-combat factors together that tightly is a mistake that harms world-building by drastically reducing the fidelity of the world.

Unoriginal

2018-07-04, 12:15 PM

Here are all the spellcaster NPCs that I could find (so far, the MM, Volo's and Mordenkainen's are covered, what's left are the modules).

Format is:

Class used as basis
Spellcaster level:
Name of the creature/statblock(CR)

Bard

4:
Bard (2)

5:
Cloud Giant Smiling One (11)

Cleric

1:
Acolyte (1/4)

2:
Kuo-toa Whip (1)

3:
Orc Eye of Gruumsh (2)

4:
Cult Fanatic (2)
Choldrith (3)

5:
Bone Naga (guardian) (4)
Priest (2)
Orc Claw of Luthic (2)

6:
Sahuagin Priestess

7:
Variant Drider (6)

9:
War Priest (9)

10:
Drow Priestess of Lolth (8)
Kuo-toa Archpriest (6)
Mummy Lord (15)

11:
Guardian Naga (10)

12:
Androsphinx (17)
Drow Inquisitor (14)

18:
Ki-rin (12)

20:
Drow Matron Mother (20)

Druid

4:
Druid (2)
Tortle Druid (2)

5:
Lizardfolk Shaman (2)

18:
Archdruid (12)

Paladin
10:
Blackguard (8)

19:
Death Knight (17)

Sorcerer

3:
Kobold Scale Sorcerer (1)

5:
Derro Savant (3)

9:
Booyahg Booyahg Booyahg (6)

13:
Skull Lord (15)

Ranger
9:
Grung Wildling (1)

Warlock

3:
Firenewt Warlock of Imix (1)
Xvart Warlock of Raxivort (1)

5:
Deathlock (4)

6:
Yuan-ti Mind whisperer (4)
Yuan-ti Nightmare Speaker (4)
Yuan-ti Pit Master (5)

7:
Neogi Master (4)

10:
Deathlock Mastermind (8)

11:
Warlock of the Archfey (4)

12:
Gloom Weaver (9)

14
Warlock of the Great Old One (6)

16:
Drow Arachnomancer (13)

17:
Warlock of the Fiend (7)

Wizard

1:
Apprentice Wizard (1/4)

2:
Hobgoblin Iron Shadow (2)

5:
Flameskull (4)
Bone Naga (spirit) (4)

7:
Hobgoblin Devastator (4)
Illusionist (3)

8:
Githyanki Gish (10)

9:
Gynosphinx (11)
Vampire Spellcaster (15)
Mage (6)
Conjurer (6)
Enchanter (5)
Transmuter (5)

10:
Drow Mage (7)
Mind Flayer Arcanist (8)
Spirit Naga (8)

11:
Morkoth (11)
Drow Favored Consort (18)

12:
Alhoon (10)
Evoker (9)
Necromancer (9)

13:
Abjurer (9)
Blue Abishai (17)

15:
Diviner (8)
Nagpa (17)

16:
Arcanaloth (12)

18:
Lich (21)
Archmage (12)
Mind Flayer Lich (22)

There is no direct relation between class level and combat potential, even among humanoids.

JackPhoenix

2018-07-04, 12:18 PM

Multiattack and Extra Attack are not the same feature. Nor are they supposed to be.

Extra Attack is a specific exception to the "when you take the attack action you make one attack" rule. It has specific considerations how it stacks (it doesn't, unless you're a fighter and then only with the fighter features) and what you can do (only make more weapon attacks or things that directly substitute for weapon attacks during the Attack action).

Multiattack is a specific action. It's not the attack action at all. It completely ignores the usual Attack action limits and requirements, substituting a whole different set of capabilities and limitations. Among those:

a) you can do things other than make another weapon attack, but only what's specified. Some can use other actions such as a dragon's fear aura ability. Others can make much more than one attack, regardless of "level."
b) you can't do things that trigger off of the Attack action (including grappling or shoving). Creatures that should have those capabilities get them built into the attacks themselves.
c) it simplifies a DM's overhead. On each turn, an enemy can do one and only one of the things listed under its "Action" block unless it specifically says otherwise. You don't have to worry about actions and bonus actions and conditionals unless you really want to. Pick one entry, resolve. Move on to the next creature. If you build everything like a PC, you lose this advantage and you dramatically increase the overhead of the DM and thus the length of the DM's turn. Or you force DMs to use fewer enemies, which has all sorts of screwy results.
D) it doesn't remove the capability of enemies to take the regular attack action and use any weapon they possess or grapple or shove. But it does mean that they can't do that as part of a multi-attack routine.

I mean... that's exactly why I said what you've quoted. Obviously, in some cases, it makes sense. In others, not so much. If it's a beast going through claw-claw-bite routine, sure. Or pseudo-TWF built into some NPCs, like Swashbuckler. But limiting the ability to use grapples or shoves as part of the attack sequence is annoying, if it's an NPC like thug who just uses Multiattack to simulate Extra Attack anyway.

PhoenixPhyre

2018-07-04, 12:23 PM

There is no direct relation between class level and combat potential, even among humanoids.

Exactly. So the whole process is fatally flawed from the get go.

You could build two NPCs starting from the same base (same class levels, race, and stats) and end up in completely different combat-potential brackets, especially for spell-casters, whose combat potential depends critically on their spell selection.

Heck, that archmage? He's only about CR 7 straight up (dCR 4-ish, oCR 10-ish). It's his spells (giving him immunity to a whole lot of spells, for one) that bump him up. If you substituted some of his spells for more blasty ones, he'd be CR 18+ (I did the calculations once...they get quite absurd) just due to his stupidly-high theoretical max damage output. Same stat block except for spells.

PhoenixPhyre

2018-07-04, 12:28 PM

I mean... that's exactly why I said what you've quoted. Obviously, in some cases, it makes sense. In others, not so much. If it's a beast going through claw-claw-bite routine, sure. Or pseudo-TWF built into some NPCs, like Swashbuckler. But limiting the ability to use grapples or shoves as part of the attack sequence is annoying, if it's an NPC like thug who just uses Multiattack to simulate Extra Attack anyway.

If you had some monsters with Extra Attack and some with Multiattack, it would get seriously confusing. Remember that NPCs are supposed to last about 3 rounds in combat. That's not enough time for complex things and the overhead gets tremendous. It also reminds the DM that these are not level N fighters.

Take the Knight NPC. He's not a level 5+ fighter, as is clear by the reaction parry ability. They could have just had them do one, bigger attack. But that causes really swingy mechanics. It's advised to give more, smaller attacks. There's nothing that actually requires that monsters wielding longswords deal 1d8 + STR damage. They could do any arbitrary amount by explicit RAW; the standard values are just a normal default tuned to make the offense/defense numbers come out right.

Unoriginal

2018-07-04, 12:39 PM

Exactly. So the whole process is fatally flawed from the get go.

You could build two NPCs starting from the same base (same class levels, race, and stats) and end up in completely different combat-potential brackets, especially for spell-casters, whose combat potential depends critically on their spell selection.

Heck, that archmage? He's only about CR 7 straight up (dCR 4-ish, oCR 10-ish). It's his spells (giving him immunity to a whole lot of spells, for one) that bump him up. If you substituted some of his spells for more blasty ones, he'd be CR 18+ (I did the calculations once...they get quite absurd) just due to his stupidly-high theoretical max damage output. Same stat block except for spells.

Indeed, indeed.

JackPhoenix

2018-07-04, 12:43 PM

If you had some monsters with Extra Attack and some with Multiattack, it would get seriously confusing. Remember that NPCs are supposed to last about 3 rounds in combat. That's not enough time for complex things and the overhead gets tremendous. It also reminds the DM that these are not level N fighters.

Take the Knight NPC. He's not a level 5+ fighter, as is clear by the reaction parry ability. They could have just had them do one, bigger attack. But that causes really swingy mechanics. It's advised to give more, smaller attacks. There's nothing that actually requires that monsters wielding longswords deal 1d8 + STR damage. They could do any arbitrary amount by explicit RAW; the standard values are just a normal default tuned to make the offense/defense numbers come out right.

Well... consistency. Generally, if there's a reason why weapon-using monster doesn't do just the weapon damage, it's spelled out in the stat block. Mostly because you know the players will want to nab the weapons if they seem more effective than their gear... we've seen it enough with threads about bladelocks wanting to wield minotaur axes or ice devil spears.

My problem with Multiattack is that, RAW, you can't substitute the attacks for diffent actions even in cases where it would make sense. A level 5 Fighter gets two attacks, and he can replace either or both with shove or grapple, as needed. A Thug can't do that, even if he otherwise can attack twice with the same weapon, just like the Figher. I'm not saying that I want Thug to copy Fighter class, but it's annoying, even if I can easily houserule it. It doesn't always make sense to do that... again, beasts who use Multiattack to bite and claw the opponent are a different matter from humanoid warriors. It is clearly apparent that the Thug's multiattack is there to simulate class ability, just like Archmage's spellcasting is (despite what He-who-shall-be-banned-if-he-posts-his-spam-again claims when he tries to make his Magic Jar bull**** work).

JoeJ

2018-07-04, 12:57 PM

It doesn't make any sense. But observing tool proficiency use doesn't provide that kind of information anyway, so it's a moot point.

So what information about an NPCs abilities are you saying that a PC could deduce from one short observation?

PhoenixPhyre

2018-07-04, 01:04 PM

Well... consistency. Generally, if there's a reason why weapon-using monster doesn't do just the weapon damage, it's spelled out in the stat block. Mostly because you know the players will want to nab the weapons if they seem more effective than their gear... we've seen it enough with threads about bladelocks wanting to wield minotaur axes or ice devil spears.

My problem with Multiattack is that, RAW, you can't substitute the attacks for diffent actions even in cases where it would make sense. A level 5 Fighter gets two attacks, and he can replace either or both with shove or grapple, as needed. A Thug can't do that, even if he otherwise can attack twice with the same weapon, just like the Figher. I'm not saying that I want Thug to copy Fighter class, but it's annoying, even if I can easily houserule it. It doesn't always make sense to do that... again, beasts who use Multiattack to bite and claw the opponent are a different matter from humanoid warriors. It is clearly apparent that the Thug's multiattack is there to simulate class ability, just like Archmage's spellcasting is (despite what He-who-shall-be-banned-if-he-posts-his-spam-again claims when he tries to make his Magic Jar bull**** work).

It's not supposed to simulate anything class-based. It has the same result, but is a completely separate feature. A fighter 5 gets one Attack action, that he can use to make any combination of attacks, grapples, and shoves (max 2). A Thug gets specific actions, one of which is to attack twice. They're separate concepts with separate, asymmetric purposes.

Of course, ad hoc house-ruling that they can substitute attacks for grapples and shoves costs exactly nothing and raises exactly no problems. Removing multiattack and replacing it with extra attack whole-sale would break things.

For example, take the gladiator NPC. He can make 3 melee attacks or 2 ranged attacks. Replacing this with Extra Attack would mean either 2 of each or 3 of each. And would allow him to use his Shield Bash and then throw twice.

Or the Bandit Captain: He can do 2 scimitar + 1 dagger (all at full damage, which wouldn't work for TWF) OR two ranged attacks with daggers. No need to track which weapons are in hands or deal with the object-interaction action economy, because specific beats general. That would go away with Extra Attack. You'd have to worry about which ones are off-hand, which is a pain.

Or the Assassin NPC. He gets two shortsword attacks. That's TWF, right? No, he gets his full ability mod to both sets of damage with no fighting style. And note he doesn't have to reapply his poison every turn (unlike a PC).

MaxWilson

2018-07-04, 01:50 PM

There's a reason why "NPCs use the same rules as PCs" was abandoned after being used in only a single edition, which was an overcomplicated mess overall.

I'm not sure which edition you're referring to, but "NPCs use the same rules as PCs" was generally (but not exclusively) true for both editions of AD&D. If it had been otherwise, if all NPCs had been using monster rules from the get-go, the term "NPC" probably wouldn't even exist.

The game works best IMO when player characters and important non-player characters follow the same rules. (Which doesn't necessarily imply that they need to be at the same power level. There's nothing wrong with making the king a 0th-level nothing.)

The "PC's and NPC's are made by different rules" is possibly my biggest (and maybe only) gripe about 5th edition.

Through all previous editions of D&D (barring 4th), and many other systems, I have always built NPCs using the same rules as player characters.

That's more of an issue with the Internet metagame than with 5E, honestly. Not only did rules for building NPCs like PCs come out first, long before the MM statblock NPCs came out, but then the 5E DMG reiterates the point that you can build NPCs using PHB rules or just make up stats like you do for monsters.

It's somewhat common to see people claiming on Internet forums that 5E NPCs are "not built like PCs," but these people are wrong in the same way that people would be wrong to claim that e.g. "human beings are not female." Some NPCs are not built like PCs, but whether you use any of those NPCs at your table is up to you.

To be fair, it is one reason I am glad for Volo's guide - I may not be a fan of the lore sections, but it gives me the info I need to create key boss monsters in the manner I prefer.

I'd like it better if there were some kind of actual correspondence between the MM stat blocks and the Volo monsters. When I first had a hobgoblin PC Sorcerer at my table (converted NPC actually, originally a captive) I just gave him martial weapons proficiency and armor proficiency (and bog-standard hobgoblin stats) plus martial advantage and called it good. In Volo's, PC hobgoblins don't get martial advantage or proficiencies at all, but they do get Saving Face. Weird. I believe bugbears are like that too. If you are trying to create key boss monsters of Volo races, it could get annoying to have the boss monsters be so different from their putative fellows. Personally I'd just ignore Volo's in that case and use the same approach as the 5E DMG rules at the back of the book, which involve just adding class levels directly onto the monster stat block.

Xetheral

2018-07-04, 02:57 PM

So what information about an NPCs abilities are you saying that a PC could deduce from one short observation?

For example, witnessing a PC-built NPC cast Mass Suggestion means you know they have at least 11 levels of Bard, Sorcerer, Warlock, or Wizard. You can infer that they have a proficiency bonus of at least +4, and at least 11 HD. You also know that they're down a sixth-ninth level slot (or their sixth level Mystic Arcanum) until they can take a Long Rest.

Witnessing a PC-built NPC smite lets one infer that they have at least two levels of Paladin or five levels of Warlock.

Witnessing a PC-built NPC take half damage as a reaction lets one infer they have at least five levels of Rogue.

And I suppose you tell them the exact abilities they use? Because "We've send a squad of guys against the renegade wizard, and he burned them all with blasts of fire" doesn't tell you anything about his real capabilities. No matter if he's build as PC or using NPC statblock. And I wouldn't expect players need to "research" every opponent they come across before they face him in battle.

If the PCs are observing directly, then yes, I'd tell the players the exact abilities being used. The players can then filter that OOC knowledge into their PC's IC knowledge of the NPCs' capabilities. Letting the abstract model "work both ways" so to speak addresses the problem that my IC descriptions can't convey as much information as the PC's senses would provide. When questioning an NPC, that's the time for the PCs to ask smart, probing questions about the shape and duration of the fire, whether it was a spout, a rolling ball, an explosion, a pillar, etc.

I run a combat-as-war game, so yes, I expect the PCs to research NPCs that they plan to engage (or whom they expect might try to engage them). Not doing legwork before an expected encounter is leaving a potential advantage on the table, and in CaW you want every advantage you can get. Do the PCs always do this research? Not at all. (I don't run a particularly deadly CAW game, so the stakes are a bit lower than they could be.) Sometimes they don't bother, sometimes time pressure means more risks need to be taken, sometimes it isn't necessary (e.g. some NPCs are so obviously out of the PC's league that non-combat solutions are required), and sometimes there simply isn't an opportunity.

But in general, in CaW, not researching your opposition is like not reconniotering a dungeon before going in. When there is no expectation of winnable combats, not knowing what you are facing ahead of time is a good way to die very quickly. (Although in my not-particularly-deadly game, there are usually additional opportunities to retreat, but that's still less advantageous than being forewarned.)

Not really. "He's a guy who can attack pretty fast" is exactly the same information you get from NPC statblock. In either case, it doesn't actually tell you anything about how dangerous that guy is, or what else can he do.

For a statblock it indeed tells you nothing. For a PC-built NPC, it provides several possible classes the NPC might have that would result in the ability to attack quickly, sets minimum floors for HP and proficiency bonus, etc. Even if the information isn't enough to identify a class uniquely, it's still useful.

How so? You've never had the PCs pick a fight with random NPCs before? If the party barbarian randomly decide to start a bar brawl, I simply decide that most people there are commoners, there may be some better brawlers who are best to be modelled by bandits or guards, and there's the local tough and bully, who's good enough to be thug. About 5 seconds of work. If I really want to, I may give them more HP or modify their stats on spot, which takes about 5 seconds more. You'll have to stat them up like PCs, with all that involve.

If there are NPCs for whom it's appropriate to have combat abilities in the bar, yes, I'll stat them up (with low detail) on the fly. But otherwise I just narrate the brawl.

And as for "entirely different abilities", unless every student is clone of each other, that will happen too. Need a "rogue"? Spy statblock got my back. Need better rogue? Assassin. Or worse rogue? I can simply add 1d6 sneak attack to a bandit statblock. Amount of work and time involved? Again, about 5 seconds. And the result is similar enough for any actuall in-game needs.

The Spy statblock has multiattack, and the 3rd level PC rogue with the same training does not. That's exactly the kind of modelling artifact I'm trying to avoid. Sure, I could tweak the Spy statblock until it matches a 3rd level rogue, but that's identical to just building it as a (low-resolution) PC.

So why do you have problem with "NPC class" abilities being different from PC class abilities? Considering you care about this while "stating NPCs as PCs", I assume you're familiar with 3.x NPC classes. They are similar to PC classes, but no sane player would choose to play one. Using a Thug statblock instead of a level 5 fighter isn't different from using a level 5 Warrior instead of level 5 fighter. "Thug" is simply a member of NPC class with d8 hit die, pack tactic, extra attack (yeah, screw multiattack limitations, I agree with that), and no other class features.

That's an interesting way of looking at it, and if the statblocks were strict subsets of PC class abilities I might indeed find them more useful. But the statblocks are not strict subsets--they give abilities and combinations of abilities that PCs can't match.

JNAProductions

2018-07-04, 02:59 PM

So you'd prefer that literally anything humanoid is entirely limited by what PC classes can do?

That seems... Restrictive.

MaxWilson

2018-07-04, 03:03 PM

For a statblock it indeed tells you nothing. For a PC-built NPC, it provides several possible classes the NPC might have that would result in the ability to attack quickly, sets minimum floors for HP and proficiency bonus, etc. Even if the information isn't enough to identify a class uniquely, it's still useful.

For example, if you know from spy reports that the Evil Lord Ruler once decapitated four peasants in one round with his greatsword, you can be reasonably confident that he doesn't also know 9th level spells.

How so? You've never had the PCs pick a fight with random NPCs before? If the party barbarian randomly decide to start a bar brawl, I simply decide that most people there are commoners, there may be some better brawlers who are best to be modelled by bandits or guards, and there's the local tough and bully, who's good enough to be thug. About 5 seconds of work. If I really want to, I may give them more HP or modify their stats on spot, which takes about 5 seconds more. You'll have to stat them up like PCs, with all that involve.

It doesn't take much effort to stat someone up like a PC. Just as I allow players to delay or elide unimportant decisions when they roll up a new character ("you can pick your languages-known later, but for now let's start the game") I can elide unimportant information about NPCs. In the bar scene I can make most of the participants 0th level nobodies (some strong laborers, others weak bums) and throw in a couple of trained guards who are first or second level fighters. Or if I think they're not trained enough to be fighters, I can make them first or second-level Warriors (members of a deficient fighter class with no class features beyond HP and armor/weapon proficiencies, plus Extra Attack at level 5 and 11).

If a player wanted to play a Warrior I'd let them, but there are no advantages relative to just playing a Fighter so they probably wouldn't. Edit: I suppose I could give them a 20% break on XP requirements for advancement, though. It's still probably much worse than a full Fighter.

JNAProductions

2018-07-04, 03:05 PM

For example, if you know from spy reports that the Evil Lord Ruler once decapitated four peasants in one round with his greatsword, you can be reasonably confident that he doesn't also know 9th level spells.

Hasted Bladesinger with GWM. Two attacks from basic action, one from Haste, one from GWM Bonus.

MaxWilson

2018-07-04, 03:14 PM

Hasted Bladesinger with GWM. Two attacks from basic action, one from Haste, one from GWM Bonus.

Mmm hmmm. And how many Bladesingers out there are going to go around wielding greatswords? It doesn't even work with Bladesong. Still fewer Bladesingers are going to invest in GWM, or Haste themselves in order to kill peasants.

You're right that he could theoretically still know 9th level spells but I'm still reasonably confident. He could be a Fighter 2/Bladelock 18 also who once used Action Surge to kill those four peasants--that is more likely than your Bladesinger scenario--but the odds of him knowing 9th level spells are much less than they were before you learned the story about the four peasants. It is useful information.

Xetheral

2018-07-04, 03:15 PM

So you'd prefer that literally anything humanoid is entirely limited by what PC classes can do?

That seems... Restrictive.

I'm fine with exceptions, but they should be available to anyone in the game world who is similarly situated. For example, if an NPC has a weird ability as a side effect of a curse, that's fine--if a PC were to be similarly cursed they would have the same ability. But I don't want the results of the curse to differ depending on whether the character in question is a PC or an NPC.

Requilac

2018-07-04, 03:20 PM

I'm not sure which edition you're referring to, but "NPCs use the same rules as PCs" was generally (but not exclusively) true for both editions of AD&D. If it had been otherwise, if all NPCs had been using monster rules from the get-go, the term "NPC" probably wouldn't even exist.

I can't speak for older editions, but when it comes to 5e, monsters are built entirely different from PCs. At least in official books. I am sure there is some homebrew out there which does otherwise.

That's more of an issue with the Internet metagame than with 5E, honestly. Not only did rules for building NPCs like PCs come out first, long before the MM statblock NPCs came out, but then the 5E DMG reiterates the point that you can build NPCs using PHB rules or just make up stats like you do for monsters.

It's somewhat common to see people claiming on Internet forums that 5E NPCs are "not built like PCs," but these people are wrong in the same way that people would be wrong to claim that e.g. "human beings are not female." Some NPCs are not built like PCs, but whether you use any of those NPCs at your table is up to you.

I am sorry if I sound agressive at all, this isn't my intention. I understand where you are coming from and have no ill feelings towards you.

The DMG does state that you build NPCs using PHB rules, but no official monster is built this way. Even the NPCs from Volo's which have the exact name of certain PC subclasses aren't built the same way as a PC. Volo's Evoker wizard doesn't have any mention of Arcane Recovery, Evocation Savant, Potent Cantrip, or Empowered Evocation, and it has a d8 hit die instead of a d6. What makes you believe that only some but not all NPCs are built like PCs? Do you have an example?

I can understand the statement that you believe that monsters and PCs in 5e should be made the same, but making the statement that they already are is not accurate.

DnDegenerates

2018-07-04, 03:31 PM

Commoner stats for most npcs. Even many hostile ones in town.

Special npcs get the named version appropriate to their role as an npc.

If they don't have special training or considerate extensive experience in any given role or job, they are a commoner.

JoeJ

2018-07-04, 03:31 PM

For example, witnessing a PC-built NPC cast Mass Suggestion means you know they have at least 11 levels of Bard, Sorcerer, Warlock, or Wizard. You can infer that they have a proficiency bonus of at least +4, and at least 11 HD. You also know that they're down a sixth-ninth level slot (or their sixth level Mystic Arcanum) until they can take a Long Rest.

Witnessing a PC-built NPC smite lets one infer that they have at least two levels of Paladin or five levels of Warlock.

Witnessing a PC-built NPC take half damage as a reaction lets one infer they have at least five levels of Rogue.

Why does it makes sense that any of those would be true in a world with millions or possibly billions of inhabitants? I can't see any reason why being really good at dodging attacks (to choose one of your examples) automatically means that one is also good at making sneak attacks and can speak the secret language of the criminal underworld.

Unoriginal

2018-07-04, 03:36 PM

For example, witnessing a PC-built NPC cast Mass Suggestion means you know they have at least 11 levels of Bard, Sorcerer, Warlock, or Wizard. You can infer that they have a proficiency bonus of at least +4, and at least 11 HD. You also know that they're down a sixth-ninth level slot (or their sixth level Mystic Arcanum) until they can take a Long Rest.

How do you witness they used Mass Suggestion?

Witnessing a PC-built NPC smite lets one infer that they have at least two levels of Paladin or five levels of Warlock.

How do you know they did smite?

Witnessing a PC-built NPC take half damage as a reaction lets one infer they have at least five levels of Rogue.

How are you going to witness a NPC taking half damage? How do you know they used a Reaction to do that?

The Spy statblock has multiattack, and the 3rd level PC rogue with the same training does not. That's exactly the kind of modelling artifact I'm trying to avoid. Sure, I could tweak the Spy statblock until it matches a 3rd level rogue, but that's identical to just building it as a (low-resolution) PC.

It's not an artifact, it's made on purpose.

MaxWilson

2018-07-04, 03:38 PM

I can't speak for older editions, but when it comes to 5e, monsters are built entirely different from PCs. At least in official books. I am sure there is some homebrew out there which does otherwise.

The 5E DMG is an "official book" last time I checked. Check towards the back, in the section on creating NPCs, and it tells you that there are two methods: build them using PHB rules, or build them like monsters. It also has a little table for stat mods for various "monster" races in case you want to do something like build a Skeleton Shadow Monk. I'm AFB or I'd give you a page number.

I am sorry if I sound agressive at all, this isn't my intention. I understand where you are coming from and have no ill feelings towards you.

No worries. I wasn't thinking of you at all when I wrote that. It's a general statement about a trend I see, not about an individual fitting that trend. I have no ill will towards you either, I just think you're incorrect.

The DMG does state that you build NPCs using PHB rules, but no official monster is built this way.

What does this even mean? Are you saying that no monster stat block in the MM is built using the PHB rules? Of course they aren't! If you are using PHB rules you don't need a MM-style stat block. Are you saying WotC hasn't published any adventures with monsters or NPCs built using PHB rules? That seems dubious to me, but also beside the point. Adventure modules aren't rulebooks. Even if it were true it would just prove that WotC adventure writers have strong preferences.

Even the NPCs from Volo's which have the exact name of certain PC subclasses aren't built the same way as a PC. Volo's Evoker wizard doesn't have any mention of Arcane Recovery, Evocation Savant, Potent Cantrip, or Empowered Evocation, and it has a d8 hit die instead of a d6. What makes you believe that only some but not all NPCs are built like PCs? Do you have an example?

You mean, from a published WotC adventure, or what?

JNAProductions

2018-07-04, 03:42 PM

It says it's possible, not that it's recommended.

Furthermore, I'm not saying "Never build NPCs like PCs". I don't think anyone is.

What IS being said is "There's little to be gained and much to lose from trying to have total PC/NPC transparency."

MaxWilson

2018-07-04, 03:48 PM

It says it's possible, not that it's recommended.

Neither approach is recommended. It just says you can do either.

I sometimes see posts from people apparently claiming otherwise, that NPCs "are not" built using PC rules. If they said "are sometimes not" built using PC rules, or "are not always", I wouldn't have an issue with it, but it seems that some people are ignorant of the fact that the PHB was not written solely for PCs.

PhoenixPhyre

2018-07-04, 03:57 PM

Apropos of this thread, I found I had done some work for various levels of a GWF (no feats) mountain dwarf champion fighter (because it was the easiest to assess).

Here are the calculated CR values (in half-CR increments) for each of a few relevant levels:

Level
oCR
dCR
CR

1
1.5
1.5
1.5

5
3
5
4

11
6
8.5
7

17
10.5
10
10

20
12
12.5
12

Make it what you will.

Requilac

2018-07-04, 04:00 PM

The 5E DMG is an "official book" last time I checked. Check towards the back, in the section on creating NPCs, and it tells you that there are two methods: build them using PHB rules, or build them like monsters. It also has a little table for stat mods for various "monster" races in case you want to do something like build a Skeleton Shadow Monk. I'm AFB or I'd give you a page number.

I know that in the DMG it says you can build monsters like players, but no monster in the MM, PHB or DMG is actually built this way.

No worries. I wasn't thinking of you at all when I wrote that. It's a general statement about a trend I see, not about an individual fitting that trend. I have no ill will towards you either, I just think you're incorrect.

Good to hear.

What does this even mean? Are you saying that no monster stat block in the MM is built using the PHB rules? Of course they aren't! If you are using PHB rules you don't need a MM-style stat block. Are you saying WotC hasn't published any adventures with monsters or NPCs built using PHB rules? That seems dubious to me, but also beside the point. Adventure modules aren't rulebooks. Even if it were true it would just prove that WotC adventure writers have strong preferences.

Your statement was that some NPCs were built like Player Characters. I countered your statement by saying that no monster or NPC in the MM, PHB, DMG, any Adventure Book (that I know of) or any other content published by WotC were made like PCs. Out of every single last fightable enemy out there which WotC has made, with the exception of some traps, not one of them are built like a PC. Literally none of them. Only Player Characters are built like Player Characters, nothing else follows the same rules or has classes like a player. The only exception to this rule is in 3rd party content and home brew. That is not my opinion, it is a completely objective observation.

You can say that you believe that some monsters or NPC's should be built like a PC, and I am saying that no monster or NPC made by WotC is actually built like a PC.

MaxWilson

2018-07-04, 04:15 PM

You can say that you believe that some monsters or NPC's should be built like a PC, and I am saying that no monster or NPC made by WotC is actually built like a PC.

Per above discussion, even if this is true (I'll take your word for it instead of looking through my adventure modules; and of course the PHB/DMG/MM don't have NPCs in them since they are rulebooks and not campaign settings) it's beside the point. It just shows that WotC adventure writers have strong preferences for how they write NPCs into adventures, same as they have strong preferences for how much treasure to award.

If the writers of Curse of Strahd choose to write Mordenkainen up as an Archmage (bleh) instead of a 20th level wizard, that reflects on them as adventure writers, but not on the 5E ruleset. It doesn't mean that it's somehow unorthodox for a given DM to run Mordenkainen as a 20th level wizard. That is squarely within the mainstream of expected 5E practice.

JackPhoenix

2018-07-04, 04:24 PM

PHB/DMG/MM don't have NPCs in them since they are rulebooks and not campaign settings

There's a whole appendix in MM named "Nonplayer Characters". It contains NPC stat blocks.

PhoenixPhyre

2018-07-04, 04:32 PM

There's a whole appendix in MM named "Nonplayer Characters". It contains NPC stat blocks.

And others in volos and mordy's guides.

Transparent dodges like that are evidence of a frail argument.

And no, MaxWilson, if something is technically allowed but not done in any published work, it's a good sign it's not mainstream. It's not wrong (because that's ill defined here), but it's unusual at least.

Unoriginal

2018-07-04, 04:44 PM

There's a whole appendix in MM named "Nonplayer Characters". It contains NPC stat blocks.

And others in volos and mordy's guides.

Transparent dodges like that are evidence of a frail argument.

Maybe I'm just crazy, but aren't, you know, all monster statblocks NPC statblocks?

And no, MaxWilson, if something is technically allowed but not done in any published work, it's a good sign it's not mainstream. It's not wrong (because that's ill defined here), but it's unusual at least.

Indeed. As an example of that, the Death Cleric and Oathbreaker Paladin are technically NPC subclasses, but WotC never used them in anything published. They just put the subclasses in the DMG, and that's all.

MaxWilson

2018-07-04, 04:50 PM

Indeed. As an example of that, the Death Cleric and Oathbreaker Paladin are technically NPC subclasses, but WotC never used them in anything published. They just put the subclasses in the DMG, and that's all.

I'm AFB but I believe the Volo's monster races are also designated primarily for NPCs as well. Perhaps I'm misremembering.

Unoriginal

2018-07-04, 04:59 PM

I'm AFB but I believe the Volo's monster races are also designated primarily for NPCs as well. Perhaps I'm misremembering.

You are, though the misremembering is understandable.

The DMG has a section on racial features for NPCs. The Volo's as a section with racial features for the traditional monster races, but they're not the same as the ones in the DMG, because the ones in the Volo's are for PCs.

For example, "Fury of the Small" is a feature for goblin PCs, but none of the goblin-as-NPC sources (be it the MM, the Volo's section on goblins, the DMG, or modules with goblins) ever mention it, while the Lizardfolk NPC racial feature in the DMG mentions a -2 to INT, which is not present in the Volo's version.

JackPhoenix

2018-07-04, 05:01 PM

Snip

Got a wrong name on that last quote, but given the circumstances, it's somewhat understandable.

The DMG has a section on racial features for NPCs. The Volo's as a section with racial features for the traditional monster races, but they're not the same as the ones in the DMG, because the ones in the Volo's are for PCs.

For example, "Fury of the Small" is a feature for goblin PCs, but none of the goblin-as-NPC sources (be it the MM, the Volo's section on goblins, the DMG, or modules with goblins) ever mention it, while the Lizardfolk NPC racial feature in the DMG mentions a -2 to INT, which is not present in the Volo's version.

Which is one of the problems I have with Volo's races. I consider being completely different from the abilities monsters of the same species have a big fault. And don't get me started on hobgoblins.

Unoriginal

2018-07-04, 05:20 PM

Got a wrong name on that last quote, but given the circumstances, it's somewhat understandable.

Thanks, and sorry.

Xetheral

2018-07-04, 05:56 PM

Why does it makes sense that any of those would be true in a world with millions or possibly billions of inhabitants? I can't see any reason why being really good at dodging attacks (to choose one of your examples) automatically means that one is also good at making sneak attacks and can speak the secret language of the criminal underworld.

It makes just as much sense for an NPC as it does for a PC. If I had a problem with a class-based abstraction, I'd be playing a classless system like GURPS.

I like both methods, for the record. But for all the reasons discussed above, I have zero interest in having PCs be modelled using a class-based system and NPCs be modelled using a classless system. Accordingly, I use the rule in the DMG that lets me build NPCs like PCs.

How do you witness they used Mass Suggestion?

Either via an ad-hoc approach to spell identification based on the PHB, the optional spell-identification rules in Xanathar's, or just by observing the casting and effects and guessing.

How do you know they did smite?

Either by being told so if the DM declares NPC actions outright, or via inferences from the in-game description if they don't.

How are you going to witness a NPC taking half damage? How do you know they used a Reaction to do that?

Again, either via explicit action declaration or in-game description. I for one make sure to declare when an antagonist uses a Reaction so that the PCs know it's spent.

It's not an artifact, it's made on purpose.

The purposefulness (or even desirability) of a modeling artifact doesn't make it stop being a modeling artifact. In this case, the mechanical differences between two identical characters, one a PC and one a statblock-style NPC, are a result of the decision to model the two characters differently. That makes it a modeling artifact, by definition.

MaxWilson

2018-07-04, 06:11 PM

You are, though the misremembering is understandable.

The DMG has a section on racial features for NPCs. The Volo's as a section with racial features for the traditional monster races, but they're not the same as the ones in the DMG, because the ones in the Volo's are for PCs.

For example, "Fury of the Small" is a feature for goblin PCs, but none of the goblin-as-NPC sources (be it the MM, the Volo's section on goblins, the DMG, or modules with goblins) ever mention it, while the Lizardfolk NPC racial feature in the DMG mentions a -2 to INT, which is not present in the Volo's version.

Thanks for checking.

JoeJ

2018-07-04, 06:12 PM

It makes just as much sense for an NPC as it does for a PC. If I had a problem with a class-based abstraction, I'd be playing a classless system like GURPS.

I like both methods, for the record. But for all the reasons discussed above, I have zero interest in having PCs be modelled using a class-based system and NPCs be modelled using a classless system. Accordingly, I use the rule in the DMG that lets me build NPCs like PCs.

As long as you're having fun you should do it that way, then. For me, building most NPCs as if they were PCs would completely destroy any sense of verisimilitude.

Unoriginal

2018-07-04, 06:22 PM

The purposefulness (or even desirability) of a modeling artifact doesn't make it stop being a modeling artifact. In this case, the mechanical differences between two identical characters, one a PC and one a statblock-style NPC, are a result of the decision to model the two characters differently. That makes it a modeling artifact, by definition.

You can't say "the characters are identical, but their mechanics are not the same, so it's bad".

If the characters are different in their mechanics/how they are modeled, they are not identical.

So no, it's only a modeling artifact if your argument that two different characters are identical is true, and it cannot be.

Thanks for checking.

You're welcome.

As long as you're having fun you should do it that way, then.

This is important. We may bicker and argue, but what's matter is that everyone is having fun in their game in the end.

Xetheral

2018-07-04, 07:17 PM

You can't say "the characters are identical, but their mechanics are not the same, so it's bad".

If the characters are different in their mechanics/how they are modeled, they are not identical.

So no, it's only a modeling artifact if your argument that two different characters are identical is true, and it cannot be.

Assume that two characters are identical in every way except mechanics. I agree that means they are not identical. My point is that if the difference in mechanics stems from the choice of modeling one character as a PC and one character as an NPC, then, by definition, the fact that the characters are not identical is a modeling artifact.

JoeJ

2018-07-04, 09:04 PM

Assume that two characters are identical in every way except mechanics. I agree that means they are not identical. My point is that if the difference in mechanics stems from the choice of modeling one character as a PC and one character as an NPC, then, by definition, the fact that the characters are not identical is a modeling artifact.

Why would somebody use different mechanics if what they wanted was two characters who are identical in every way?

PhoenixPhyre

2018-07-04, 09:11 PM

Xetheral

2018-07-04, 10:18 PM

Why would somebody use different mechanics if what they wanted was two characters who are identical in every way?

And I'm not sure why one would want two identical characters in the first place. The idea boggles my mind--no two creatures are identical.

It's an illustration of where the modeling artifact comes from and an explanation of why I'm using the term correctly (which Unoriginal has disputed). It's not meant to be a literal example.

Pex

2018-07-04, 11:05 PM

And I'm not sure why one would want two identical characters in the first place. The idea boggles my mind--no two creatures are identical.

Saves time and work. Copy and paste into your game as many as needed.

MaxWilson

2018-07-04, 11:09 PM

And I'm not sure why one would want two identical characters in the first place. The idea boggles my mind--no two creatures are identical.

Judging by upthread, there are several people who want identical characters--or at least, they consider NPC archetypes in the MM to be fully equivalent to actual NPCs, and get snippy if you try to draw a distinction or point out that specific NPCs belong in specific adventures/settings, not the rulebooks. (Therefore, it's not surprising if there are no NPCs at all in the DMG/PHB/MM.)

JoeJ

2018-07-05, 01:12 AM

Therefore, it's not surprising if there are no NPCs at all in the DMG/PHB/MM.

The MM has nothing but NPCs.

Kadesh

2018-07-05, 07:05 AM

Judging by upthread, there are several people who want identical characters--or at least, they consider NPC archetypes in the MM to be fully equivalent to actual NPCs, and get snippy if you try to draw a distinction or point out that specific NPCs belong in specific adventures/settings, not the rulebooks. (Therefore, it's not surprising if there are no NPCs at all in the DMG/PHB/MM.)
Here is how I have literally just built an NPC boss fight in my games.

Take Champion. Give Half Silver Dragon template. Replace Greatsword with Halberd, give every attack the Sweeping Attack effect from Battlemaster. 4 uses of Magic Missile (5th level) per day. Unlimited use of Erupting Earth (1st level) when he uses Multiattack with Halberd. He's pretty charismatic as a King as well, so his DC's aren't bad, so lets say he's a CR13-14 encounter, so maybe DC16. 1/day can Gate in an Adult Copper Dragon.

Exactly mirrored abilkties and flavour. No stress. No careful calculation of Skill Points. No feat balancing needed. Just purely stating exactly what each character can do.

GreyBlack

2018-07-05, 07:32 AM

Quick answer: ask your DM

Long answer: it all depends on setting. With the exception of maybe 5-10 NPC's who are explicitly at Eldritch Horror status in my personal campaign setting (and, for all intents and purposes, they're actually mortal gods on the Material Plane), there are no NPC's above 8th level. If there are 2 wizards walking into a village for the first time, the majority of there villagers have seen 2 wizards in their lifetime. Maybe the old crackpot at the edge of town has seen 3, and no one has ever seen a 2nd level spell ever. The regional king maybe keeps 2 Clerics on hand who can cast level 3 spells, and they're paid extraordinarily well for their services.

This is not the case in the base game, where it's more or less assumed that your character by level 5 will have a magic item and there are at least 50 people above level 10. Certain magic items are readily available for purchase.

So, to me, it all depends the flavor of the campaign. Personally, I like low magic settings there the PC's are literally the last bulwarks against the forces of the hells, but your mileage may vary.

MaxWilson

2018-07-05, 09:35 AM

Here is how I have literally just built an NPC boss fight in my games.

Take Champion. Give Half Silver Dragon template. Replace Greatsword with Halberd, give every attack the Sweeping Attack effect from Battlemaster. 4 uses of Magic Missile (5th level) per day. Unlimited use of Erupting Earth (1st level) when he uses Multiattack with Halberd. He's pretty charismatic as a King as well, so his DC's aren't bad, so lets say he's a CR13-14 encounter, so maybe DC16. 1/day can Gate in an Adult Copper Dragon.

Exactly mirrored abilkties and flavour. No stress. No careful calculation of Skill Points. No feat balancing needed. Just purely stating exactly what each character can do.

And here's how I build a similar NPC:

Lord Kadesh, Scro Eldritch Knight 12 (we'll use the Orc template from Volo's, but with no penalty to Intelligence). Give him GWM, Lucky, and Str 20/Dex 12/Con 15/Int 13/Wis 12/Cha 13. GWM and Lucky feats. Shield spell, Absorb Elements, Blur, Find Familiar, Darkness. [text on personality goes here but isn't relevant to this thread]

No stress. No careful calculations of Skill Points.

It isn't hard. What's hard is generally giving an NPC goals, personality, and alliances, but coming up with basic stats is not.

Requilac

2018-07-05, 09:39 AM

Per above discussion, even if this is true (I'll take your word for it instead of looking through my adventure modules; and of course the PHB/DMG/MM don't have NPCs in them since they are rulebooks and not campaign settings) it's beside the point. It just shows that WotC adventure writers have strong preferences for how they write NPCs into adventures, same as they have strong preferences for how much treasure to award.

If the writers of Curse of Strahd choose to write Mordenkainen up as an Archmage (bleh) instead of a 20th level wizard, that reflects on them as adventure writers, but not on the 5E ruleset. It doesn't mean that it's somehow unorthodox for a given DM to run Mordenkainen as a 20th level wizard. That is squarely within the mainstream of expected 5E practice.

Judging by upthread, there are several people who want identical characters--or at least, they consider NPC archetypes in the MM to be fully equivalent to actual NPCs, and get snippy if you try to draw a distinction or point out that specific NPCs belong in specific adventures/settings, not the rulebooks. (Therefore, it's not surprising if there are no NPCs at all in the DMG/PHB/MM.)

No published adventure book has a monster/NPC built like a character either, even those specific NPCs which belong in specifics adventures/setting and not in rule books.

And that is not besides the point. My entire point was that no monster or NPC (including Blackstaff from LMoP, The Elemental Princes from PotA, Strahd, Acererak from ToA and Mordenkainen from CoS) published by WotC is built like a PC. You claimed that some monsters/npcs were built that way, and I was providing evidence that no monster was built that way. I am not sure what you think my claim was, but that was my claim. My claim is that WotC of the coast has never produced a monster/NPC which is built like a PC for D&D 5e.

You are correct that a DM could make a monster/NPC following PC rules, and there is guidance for it, but WotC has never done such a thing. I never claimed that a DM could not use PC rules to make a monster/NPC, that would clearly be a false claim because of information in the DMG, my claim was that WotC has never made such an NPC. If you want to make an NPC exactly like a PC, go knock yourself out. But it is obviously not in "the mainstream" of D&D 5e, it is just possible within 5e, even if WotC has never done it. Does that make sense?

MaxWilson

2018-07-06, 03:06 AM

And that is not besides the point. My entire point was that no monster or NPC (including Blackstaff from LMoP, The Elemental Princes from PotA, Strahd, Acererak from ToA and Mordenkainen from CoS) published by WotC is built like a PC. You claimed that some monsters/npcs were built that way, and I was providing evidence that no monster was built that way. I am not sure what you think my claim was, but that was my claim. My claim is that WotC of the coast has never produced a monster/NPC which is built like a PC for D&D 5e.

Thanks for clarifying your claim. Can you see now why it is not a a valid rebuttal to my previous claim, that some 5E NPCs are built with PHB rules and some are built with MM-style statblocks?

Kadesh

2018-07-06, 04:16 AM

Thanks for clarifying your claim. Can you see now why it is not a a valid rebuttal to my previous claim, that some 5E NPCs are built with PHB rules and some are built with MM-style statblocks?

I think the only point you've proved is an unwillingness to accept you are wrong and are instead willing to grasp at straws.

No NPC is built using the character profile. No NPC has a Character sheet. Why is the Character Shet much more in delth than the stat block for an NPC? Like you said, stats are the easy thing. You've not built that Eldritch Knight the same as a PC, you've ignored the Int penalty and some spells, and cut corners on irrelevant things, invalidating your own argument.

By your own admittance, there was no careful counting involved, there was simply 'ima place this here to get a basic idea of what this character can do' : and 5e have made that the core concept of how to run and build NPC's.

MeeposFire

2018-07-06, 05:10 AM

I would disagree with the idea that AD&D and older versions of D&D npcs were built like characters. I would grant you that important NPCs would often be built that way but most of them would be built as 0 level humans which is something that player characters do not do. Further if you look if you look at the various monster manuals, compendiums, or in the monster section of the Rules Cyclopedia you will find examples of player character race stat blocks and they are done up as monsters using the monster making rules not the class rules.

IN many ways this is really similar to how 5e currently handles NPCs with the only difference being that these older versions of D&D had a higher frequency of using the PC rules in making notable PCs but I would say that there is an important caveat in that making characters back then was exceedingly simple so using PC rules was not much more onerous than using the monster rules.

I remember back when I used to DM 3e that I used to put all sorts of work into my NPCs to make them mechanically viable and interesting. That was the biggest waste of time and effort I have ever done in the game. Players appreciate an NPC with personality and one that makes for a fun fight but they do not know (or typcially care) that you spent a whole bunch of time giving them all these classes, prcs, skills, feats when they get killed in on egaming session. So not worth it.

Max does show one thing I really do like about 5e when it comes to making NPCs using PC type rules and that is it IS much easier to buid one up quickly compared to say 3e, however they do have similar issues as 4e NPCs built like PCs in that they tend to be nasty offensively but really weak defensively. This means that I often use the PC rules for NPCs that I am not planning for the PCs to fight. If an NPC is planned to be part of an encounter as an enemy then they tend to be built like monsters since that makes it easier for them to fit in (also means that I can often just redress an existing stat block for Max's example I might use a human warrior type stat block and give them a few spells to show them being an eldritch knight type character) but if the character is planned to be potnetial help to the players then statting them up as a PC might be appropriate.

Using the PC stats for NPCs is a nice option to have but I am really glad that NPCs and monsters generally do not use the same rules as PCs.

JackPhoenix

2018-07-06, 06:05 AM

Requilac

2018-07-06, 07:10 AM

Thanks for clarifying your claim. Can you see now why it is not a a valid rebuttal to my previous claim, that some 5E NPCs are built with PHB rules and some are built with MM-style statblocks?

Just for clarification, your current claim is that it is possible to build an NPC using PC rules and that some people do* just that, it’s just that WotC hasn’t done it at all yet, right? If that is what you are trying to say, then I would agree with you. You are right if you would say that some DMs or Third Party Companies make NPCs built like PCs. But saying that WotC does that is false. Do I understand you correctly. If I do, then this debate is over and this case is closed.

*Meeposfire has admitted to doing so

Unoriginal

2018-07-06, 08:12 AM

Tanarii

2018-07-06, 08:47 AM

My biggest issue with using PC classes for NPCs is that the PC class HPs are just not designed to handle other PCs hammering on them for a few rounds.

So if you do that you need to make a much higher level PC (which brings other problems) or boost their HPs... At which point you're no longer using PC rules.
Similar problems on the offensive side. All class levels are roughly equal from the perspective of a team of PCs working together over a series of encounters in an adventuring day. Not even close for a single encounter, especially nova spell casters with really punchy spells.

Meanwhile, offensive CR accounts for big nova spells over three rounds in a single combat.

So on top of designing your NPCs as PCs, if they're potentially going to be PC opponents and you want to appropriately telegraph difficulty (CaW game) or directly set level appropriate encounters (CaS game) you need to calculate their CR and encounter difficulty anyway.

Not knowing how difficult the opponents are going to be, roughly, means you can't run either kind of game very well.

Xetheral

2018-07-06, 09:34 AM

Similar problems on the offensive side. All class levels are roughly equal from the perspective of a team of PCs working together over a series of encounters in an adventuring day. Not even close for a single encounter, especially nova spell casters with really punchy spells.

Meanwhile, offensive CR accounts for big nova spells over three rounds in a single combat.

So on top of designing your NPCs as PCs, if they're potentially going to be PC opponents and you want to appropriately telegraph difficulty (CaW game) or directly set level appropriate encounters (CaS game) you need to calculate their CR and encounter difficulty anyway.

Not knowing how difficult the opponents are going to be, roughly, means you can't run either kind of game very well.

I disagree that one needs to know CR and DMG Encounter Difficulty to be able to telegraph how challenging an opponent may be. If the PCs can infer (from observations or research) the approximate level and class(es) of an opponent, that's enough to provide a direct estimate of how dangerous the NPC might be to fight. Knowing that a particular NPC is rated by the DMG as an easy/moderate/hard/deadly encounter doesn't provide much additional information.

Also, in a CAW game the CR and DMG Encounter Difficulty of a particular NPC is likely to change based on the actions of the PCs (e.g. attacking an NPC wizard after paying him to cast a lot of spells for you that day, or attacking an NPC fighter when she's already low on HP or not wearing armor). Knowing (via telegraphing) that an NPC is usually CR X and Encounter Difficulty Y doesn't help the players determine how challenging the NPC would be to defeat in a particular circumstance. Knowing instead that the NPC is a level X-Y Fighter and probably doesn't have any other classes gives the players a very good idea of how changing the circumstances might change the challenge level.

It's entirely possible for the PCs to have a good idea how difficult (in the common sense) an opponent may be without the players (or the DM) knowing the CR or DMG Encounter Difficulty.

JackPhoenix

2018-07-06, 10:26 AM

I disagree that one needs to know CR and DMG Encounter Difficulty to be able to telegraph how challenging an opponent may be. If the PCs can infer (from observations or research) the approximate level and class(es) of an opponent, that's enough to provide a direct estimate of how dangerous the NPC might be to fight. Knowing that a particular NPC is rated by the DMG as an easy/moderate/hard/deadly encounter doesn't provide much additional information.

So you assume that either the players know only CR and encounter difficulty, or only level and class? Of course the later provides more informations, because a) you give 2 pieces of information instead of one (CR and encounter difficulty boils down to the same thing, and pretty much the same thing as level, class is a bonus on top) and b) you limit the possible pool of powers by sticking to classes.

You can just as well observe or research the power of an opponent built like NPC, and get about the same result.

I consider giving the players either information (CR or class and level) a bad kind of metagaming.

"You've faced zombies before, but this one moves with more purpose and coordination, also it wears plate armor and greatsword, but it's still easier opponent than evil knight you've fought before" is enough information, without requiring "You see CR2 zombie/level 3 zombie fighter. It's more dangerous than normal CR 1/4 zombies/level 1 zombie fighters, but you've defeated the evil CR 3 Knight/CR5 human fighter with plate armor and greatsword without much trouble, so it shouldn't be much of a challenge".

Also, in a CAW game the CR and DMG Encounter Difficulty of a particular NPC is likely to change based on the actions of the PCs (e.g. attacking an NPC wizard after paying him to cast a lot of spells for you that day, or attacking an NPC fighter when she's already low on HP or not wearing armor). Knowing (via telegraphing) that an NPC is usually CR X and Encounter Difficulty Y doesn't help the players determine how challenging the NPC would be to defeat in a particular circumstance. Knowing instead that the NPC is a level X-Y Fighter and probably doesn't have any other classes gives the players a very good idea of how changing the circumstances might change the challenge level.

Bad comparisons. An spellcaster NPC with wasted spells or half-dead warrior NPC without equipment is easier target regardless if it's built using class levels or NPC rules. I don't need the exact numbers to figure that out.

It's entirely possible for the PCs to have a good idea how difficult (in the common sense) an opponent may be without the players (or the DM) knowing the CR or DMG Encounter Difficulty.

True, but it's just as possible for the PCs to have a good idea how difficult an opponent may be without the players knowing the opponent's character class, because he doesn't have any. Even with no CR or class levels in real life, I know that professional soldier would be more dangerous than 80 years old grandma, or that I'd have more chance facing angry wolf than angry bear (or angry tyrannosaurus, if there was such thing).

Tanarii

2018-07-06, 10:36 AM

MaxWilson

2018-07-06, 11:28 AM

No NPC is built using the character profile. No NPC has a Character sheet. Why is the Character Shet much more in delth than the stat block for an NPC? Like you said, stats are the easy thing. You've not built that Eldritch Knight the same as a PC, you've ignored the Int penalty and some spells, and cut corners on irrelevant things, invalidating your own argument.

By your own admittance, there was no careful counting involved, there was simply 'ima place this here to get a basic idea of what this character can do' : and 5e have made that the core concept of how to run and build NPC's.

I didn't ignore any Int penalty (Scro don't have an Int penalty) and ignoring some spells/languages/etc. is done for PCs too. ("We'll pick your languages later. For now you can leave them wildcards.")

You're falling into the classic trap that I pointed out originally: people who claim that MM stat blocks are easier/simpler than PHB rules are forgetting that you can ALWAYS delay or elide resolution of irrelevant details, for both PCs and NPCs. You don't have to decide what the ogre's mother's name is unless and until the PCs capture him and start asking. In many cases you don't even need to determine the ogre's OWN name.

P.S. Character sheets? Hah! Waste of space. A character can often fit on a single line of a note book page. I've scrawled plenty of characters on the back of a napkin.

Kadesh

2018-07-06, 12:28 PM

I didn't ignore any Int penalty (Scro don't have an Int penalty) and ignoring some spells/languages/etc. is done for PCs too. ("We'll pick your languages later. For now you can leave them wildcards.")

You're falling into the classic trap that I pointed out originally: people who claim that MM stat blocks are easier/simpler than PHB rules are forgetting that you can ALWAYS delay or elide resolution of irrelevant details, for both PCs and NPCs. You don't have to decide what the ogre's mother's name is unless and until the PCs capture him and start asking. In many cases you don't even need to determine the ogre's OWN name.

P.S. Character sheets? Hah! Waste of space. A character can often fit on a single line of a note book page. I've scrawled plenty of characters on the back of a napkin.

Can you clarify your position, because you're very adept at moving goalposts. Because what I'm seeing is the fact that you're not following character creation rules and cutting corners from the process in order to prove the point that it's easy to create a character using PC rules. That's like saying beating Usain Bolt in the 100m is easy because you can traverse the same distance in a Ferrari.

Either way, have fun with whatever it is you wish to achieve with this discussion.

MaxWilson

2018-07-06, 12:36 PM

Can you clarify your position, because you're very adept at moving goalposts. Because what I'm seeing is the fact that you're not following character creation rules and cutting corners from the process in order to prove the point that it's easy to create a character using PC rules. That's like saying beating Usain Bolt in the 100m is easy because you can traverse the same distance in a Ferrari.

Either way, have fun with whatever it is you wish to achieve with this discussion.

If you think repeating myself constitutes "moving goalposts", you need to go back and read the thread. For example, from page 3 (my second post in this thread):

For example, if you know from spy reports that the Evil Lord Ruler once decapitated four peasants in one round with his greatsword, you can be reasonably confident that he doesn't also know 9th level spells.

It doesn't take much effort to stat someone up like a PC. Just as I allow players to delay or elide unimportant decisions when they roll up a new character ("you can pick your languages-known later, but for now let's start the game") I can elide unimportant information about NPCs. In the bar scene I can make most of the participants 0th level nobodies (some strong laborers, others weak bums) and throw in a couple of trained guards who are first or second level fighters. Or if I think they're not trained enough to be fighters, I can make them first or second-level Warriors (members of a deficient fighter class with no class features beyond HP and armor/weapon proficiencies, plus Extra Attack at level 5 and 11).

If a player wanted to play a Warrior I'd let them, but there are no advantages relative to just playing a Fighter so they probably wouldn't. Edit: I suppose I could give them a 20% break on XP requirements for advancement, though. It's still probably much worse than a full Fighter.

Furthermore, I think you misunderstand what "moving goalposts" means. "Goalposts" are requests for proof of claims offered by another person. If I asked you for proof of claim XYZ, and you gave it to me, and I said that wasn't good enough, that would be moving goalposts. Instead I'm offering an opinion (quoted above), and other people take issue with it claiming ABC, and I disagree and point out that that's changing the subject... all I'm doing is resisting someone else's attempt to move goalposts. In this case, your attempt. You claimed something, I rebutted it with facts. Now you're trying to move the goalposts.

Kadesh

2018-07-06, 01:46 PM

I like how your answer to this discussion is "so I can metagame better".

Xetheral

2018-07-06, 02:04 PM

MaxWilson

2018-07-06, 02:51 PM

But for you, e DM, to properly telegraph the danger to the PCs, which is a key component of CaW style, you have to have an idea of how dangerous the opponent is. Your options are use Encoutner difficulty for a straight up fight, eyeball it, or do not telegraph and let the players draw their own conclusions. The latter is basically ignore the requirement to telegraph and putting it on the players completely.

Eyeballing, or even better telegraphing specific capabilities and behaviors, will give better results than merely telegraphing CR.

What's more dangerous, a Tyrannosaurus Rex (CR 8) that charges the nearest PC and attacks whatever gets in its face until it's dead, or a Tyrannosaurus Rex (CR 8) lying in the grass that suddenly stands and charges for the softest-looking target, snaps it up in its jaws (grapple + restraint), and then withdraws at its best movement speed (50', 100' if dashing) to gnaw on its meal in peace and quiet? The second version is far more likely to kill at least one PC, but they have the exact same stats and CR.

So, if this land is full of the latter type of Tyrannosaurs, not only do you need to let the players know that these things kill and eat wolves, lions and elephants and are as strong as giants, but you also let them know that T-Rexes are fast and surprisingly cunning at picking off wolves who wander too far from the pack. CR isn't nearly enough information for good telegraphing.

MrStabby

2018-07-07, 08:08 PM

But for you, e DM, to properly telegraph the danger to the PCs, which is a key component of CaW style, you have to have an idea of how dangerous the opponent is. Your options are use Encoutner difficulty for a straight up fight, eyeball it, or do not telegraph and let the players draw their own conclusions. The latter is basically ignore the requirement to telegraph and putting it on the players completely.

I think telegraphing is important - but as pointed out CR is a poor tool. As important is telegraphing (or otherwise) what the risks are, not just how severe. Ensuring that there is substance there for a history check about how lord Xanthus defeated someone is a duel by poisoning them the night before gives information on their character, their preferred damage types, possible skill level but also for smart PCs suggesting some skills to enter the home of their opponent unobserved. If this NPC is built in a manner more closely guided by NPC classes it does give a framework to guess other abilities.

Tanarii

2018-07-07, 09:34 PM

I think telegraphing is important - but as pointed out CR is a poor tool.
Except encounter difficulty isn't really a poor tool. Its certainly better than eyeballing it.

MaxWilson

2018-07-08, 09:35 AM

Except encounter difficulty isn't really a poor tool. Its certainly better than eyeballing it.

Even in purely CAS terms, computing encounter difficulty is inferior to computing the 3/2 power of both PC and monster numbers. The "difficulty multipliers" seem to be WotC's way of roughly accounting for Lanchester's laws (see https://en.wikipedia.org/wiki/Lanchester%27s_laws, note that "In modern warfare, to take into account that to some extent both linear and the square apply often, an exponent of 1.5 is used," and then look at the XP multiplier tables and notice that result in scaling with give you approximately the 1.5 power) and in homogenous combats they give answers roughly congruent with Lanchester's laws. But when it comes to heterogenous combats, like one beholder and twenty hobgoblins, the XP multipliers overestimate how much the hobgoblins help out the beholder. Instead of taking the 3/2 power of the raw numbers, it's more accurate to take the 3/2 power of the XP values (XP values are a proxy for raw monster power: the product of offensive and defensive power). The encounter building rules in Xanathar's do not have this flaw and lead to DMs more naturally using heterogenous groups of monsters, but the Xanathar's approach doesn't give you a number that you can plug back into the daily XP budget.

So you can be more accurate by tossing out the encounter values and computing things directly, but notice that this rests on a sandy foundation: it assumes that CR and XP values are a good proxy for monster power. This is true for simple bruiser monsters that will just make attack rolls until they're dead, and false for tricksy monsters like Intellect Devourers and Banshees that can do sneaky things like phase through the walls, hit vulnerable PCs with heavy attacks, and phase back out after taking minimal damage. Ditto any monsters with complex spellcasting capabilities.

Bottom line: WotC's encounter difficulty is a good way of precisely calculating how homogenous groups of monsters fighting stupidly will do against the PCs in a cage match. The extent to which that is superior to eyeballing depends primarily on whether you're running a CAS game or a CAW game: how like a cage match are your typical fights? If the pure mechanics of rolling attacks and damage are the main thing that matters, instead of tactics, then WotC encounter difficulty is going to be a good proxy for actual difficulty, otherwise you need to eyeball it.