Fixing Knowledge checks made to identify creatures

[TuggyNE]

In a couple of threads lately I've been discussing the problem of Knowledge checks made to identify creatures — especially the absurdly well-known ones such as elves, dwarves, and humans — that are way too high, or are wildly inconsistent. An old WotC forum project fixes many of the inconsistencies, but then introduces new ones: DC 30 (!) elf lore reveals the existence of drow and other variants, such as half-elves; those variants are each DC 12 to identify. Mules are DC 11 to identify. And so on.

Therefore, I'm looking to brainstorm some more detailed guidelines for identification DCs that make sense for educated NPCs, uneducated PCs, and educated PCs. (Ideally also ignorant NPCs.)

So far, I've thought of the following:

• Each type has a base DC, perhaps 9+CR for most types and 4+CR for humanoids, animals, and giants, or perhaps a bit more nuanced
• If more than one variety of similar creature exists, a basic identification can be made at the DC of a weaker variety (e.g. red dragons, most demons, most devils, most angels, etc; creatures advancing by class are identified based on their lowest possible CR); this identification reveals only those traits, features, and abilities that the weaker variety has in common with the stronger
• Some creatures are more widely recognizable, and should probably have lower DCs listed in specific exception to their type rule (whether by -2, -4, or -5 I'm not sure)
• Bare identification DCs are calculated with an eye to (low-level) characters of Core PC races for the most part
• Characters from unusual backgrounds may have considerable variation in ease of identification; a moose may be DC 6 for those from one area, DC 10 for those from another, and DC 12 for yet another; however, this is not based on where the creature currently is, but the familiarity the would-be identifier has with it
• Getting a single bit of additional information about a creature is a +2 increase
• An untrained Knowledge check can be made to identify a creature, but circumstance penalties are likely if it's rare or unknown in the character's home

It's probably best to work upward from the simplest DCs to the more complex ones, because it's easy to guess that most people should know what a mule is (so the DC has to be no more than 10), and so on; then, using comparisons, work up to the big stuff.

Difficulty	DC	Example	Explanation
Very easy	0	Own race*
Easy	5	A mule, in a country that uses mules	Nearly everyone knows this
Average	10	A dwarf on first sight, based on stories	Most people are more or less aware of this
Tough	15	A dire bear, as opposed to its lesser cousins	Scholars, and well-educated laypeople, will know this
Challenging	20	A couatl	Most scholars will know this much
Formidable	25	A marilith	A specialist in the field is likely to be able to remember this
Heroic	30	A mithral golem	Only a specialist would be at all likely to figure this out
Nearly impossible	40	A demilich	Extensive research and a great deal of specialized study is necessary to discover this

*Yes, I'm aware this makes the old "raised by wolves" stunt nearly impossible (-5 circumstance to Knowledge (local) from bizarre childhood, -5 circumstance to self-identify from being lied to, -1 Int mod), but that's too bad.

[Mnemnosyne]

One issue that immediately occurs to me is that of regional differences in how common knowledge may be. For instance, in a desert like Calimshan, camels are probably reasonably well known and common as riding animals and beasts of burden. In an area like Icewind Dale, on the other hand, it would probably take a scholar to identify a camel.

So, some sort of rule based on the creature's environment should probably factor in, such as increasing the DC by +5 to +10 or even more based on how alien the environment is to the individual in question. For instance, a Thoqqua is a low CR monster native to the Elemental Plane of Fire that also commonly inhabits the Elemental Plane of Earth, so on those planes the DC to identify them may be pretty low for the average native. But on the Prime Material they're way rarer and could rate a +10 DC increase to identify, and might rate even higher increases on planes even more unusual for them to be on.

[TuggyNE]

One issue that immediately occurs to me is that of regional differences in how common knowledge may be. For instance, in a desert like Calimshan, camels are probably reasonably well known and common as riding animals and beasts of burden. In an area like Icewind Dale, on the other hand, it would probably take a scholar to identify a camel.

So, some sort of rule based on the creature's environment should probably factor in, such as increasing the DC by +5 to +10 or even more based on how alien the environment is to the individual in question. For instance, a Thoqqua is a low CR monster native to the Elemental Plane of Fire that also commonly inhabits the Elemental Plane of Earth, so on those planes the DC to identify them may be pretty low for the average native. But on the Prime Material they're way rarer and could rate a +10 DC increase to identify, and might rate even higher increases on planes even more unusual for them to be on.

A good point, although I don't like to be throwing around +10 modifiers willy-nilly as a general rule. I think I'll edit in something like that, yeah.

On the other hand, if enough general modifiers can be found, we might be able to get rid of HD/CR-based scaling, which would cause me to jump up and down in excitement.

Also, I suspect a sensible rule of thumb is to assume the viewpoint of low-level Core PC races for the most part, allowing modifiers for other viewpoints as suitable.

[mattie_p]

I like having different scales based on type.

Humanoid and animal type base DC is 5, giant, dragon, and monstrous humanoid base DC is 10, magical beast, construct, and plant base DC is 15, aberration and outsider base DC is 20, etc. These are suggestions, feel free to discuss.

No scaling by hit dice. A wyrmling red dragon is basically the same as a great wyrm red dragon - no reason why they should be more difficult to ID as they age. If anything, they should be easier.

Conditional modifiers based on favored terrain type. If the local terrain is plains, add +2 DC to hills, +4 DC for mountains and underground. +2 DC for forest.

The difficulty is in how to apply the base DC and conditional modifiers in a systematic way. I know that is what we are attempting to resolve here, but I fear it will be more difficult than we expect. I usually just handwave it, unfortunately.

[sonofzeal]

I like having different scales based on type.

Humanoid and animal type base DC is 5, giant, dragon, and monstrous humanoid base DC is 10, magical beast, construct, and plant base DC is 15, aberration and outsider base DC is 20, etc. These are suggestions, feel free to discuss.

Sounds good!

No scaling by hit dice. A wyrmling red dragon is basically the same as a great wyrm red dragon - no reason why they should be more difficult to ID as they age. If anything, they should be easier.

I'm torn here. There's certainly some sense to higher-HD creatures being more exotic and hence harder to identify. Honestly, it's only Dragons for which it's usually a problem. Maybe a caveat that for Dragons you use the smallest form, and for everything else you use pre-Advancement HD?

Conditional modifiers based on favored terrain type. If the local terrain is plains, add +2 DC to hills, +4 DC for mountains and underground. +2 DC for forest.

Bad idea. Too complicated, requires extra information to be stored (environment can usually be discarded when storing monster info concisely), and requires an additional table for quick reference. This would bog the game down unjustifiably.

[mattie_p]

I'm torn here. There's certainly some sense to higher-HD creatures being more exotic and hence harder to identify. Honestly, it's only Dragons for which it's usually a problem. Maybe a caveat that for Dragons you use the smallest form, and for everything else you use pre-Advancement HD?

I could be flexible here. But then you have seaside commoners who can identify the fish they catch for eating but have no idea what kind of gargantuan 12 HD creature they catch for oil. Would need to fix that somehow.

Bad idea. Too complicated, requires extra information to be stored (environment can usually be discarded when storing monster info concisely), and requires an additional table for quick reference. This would bog the game down unjustifiably.

I agree, I'd love to simplify it somehow. Honestly, the easiest way would have been if the designers included a line in each statblock on the DC to identify. As they have not, I see few alternatives, otherwise, as mentioned above, icewind dale barbarians can identify camels. But if someone else comes up with a creative solution I'd be willing to entertain it.

[Mnemnosyne]

I think perhaps the biggest problem with knowledge checks to identify creatures is the untrained rule that prohibits you from succeeding on a knowledge check of DC 11 or higher if you're untrained in the skill.

Removing or adjusting that rule could by itself go a long way toward fixing this; circumstance bonuses could then kick in for a lot of things, allowing people to identify things that are common in their area or commonly used, without having to give them the ability to identify unusual or non local things.

The seaside commoners would then get a significant circumstance bonus to identify the whale because of their familiarity with it, but they wouldn't get a bonus to a rare sea monster.

As for things like dragons, it seems reasonable to me to handle the scaling thusly: if you can identify the youngest age category, you know the basic abilities, but you have to identify an older age category if you want to know things that the dragon can't do until those age categories. Say you're fighting a great wyrm blue dragon, but you only pass the knowledge check to identify a young one. You know basics blue dragon stuff, but you don't know about it's spell casting, its DR, its SR, or most of its spell-like abilities.

[Doug Lampert]

As for things like dragons, it seems reasonable to me to handle the scaling thusly: if you can identify the youngest age category, you know the basic abilities, but you have to identify an older age category if you want to know things that the dragon can't do until those age categories. Say you're fighting a great wyrm blue dragon, but you only pass the knowledge check to identify a young one. You know basics blue dragon stuff, but you don't know about it's spell casting, its DR, its SR, or most of its spell-like abilities.

That brings up another problem. By rule you get the ID at some DC and "a bit of useful information about that monster. For every 5 points by which your check result exceeds the DC, you recall another piece of useful information."

So say a blue dragon is DC 15 and the useful peice of information is that if can fly (duh). Then at DC 20 you get that it breaths lighning (how obscure is that). At DC 25 you get that it can burrow (utterly vital if you're fighting one and it's playing to its strengths). At DC 30 you get that it has exceptionally sharp senses (also important since you effectively can't hide from an equal CR dragon). At DC 35 you get that it's fast (so you can't run away, clearly useful information). At DC 40 you get that they live in deserts and can create and destroy water (first spell-like, that's actually two bits of info but they're related). At DC 45 you get that they are highly intelligent and can talk. At DC 50 you get that they are LE (that's important once you know they can talk). At DC 55 you get that some of them are spell-casters.

Is there something wrong with that list of abilities? It strikes me as a pretty straightforward listing in the rough order of importance, and we're up to EPIC skill checks and still don't know about frightful presence or SR or DR or most of their spell-likes or that the big ones get vast numbers of attacks in melee or....

This is just silly. 1 skill rank is speaking a language fluently, it's the difference between unskilled labor at 1 sp a day and a skilled professional earning about ten times as much. It's enough to turn someone who barely made the Olympic team into a gold medalist in anything where althletics gives a performance.

One skill rank is a BIG DEAL. Yet a genious with 40 skill ranks in the appropriate skill probably does not know that blue dragons sometimes have SR or that they have frightful presence or that they have DR. Or if you don't like the order of my list change it up and see if it makes any more sense not knowing that they can burrow, or that they can talk, or that they are evil, or whatever.

Crap. Beat the DC by 10 and you should bloody well be able to simply read the monster manual listing for the creature. DM customization will assure that doesn't give you absolutely everything, but the +5 to DC for a single peice of useful information is absurd.

[mattie_p]

* Lots of awesome *

Crap. Beat the DC by 10 and you should bloody well be able to simply read the monster manual listing for the creature. DM customization will assure that doesn't give you absolutely everything, but the +5 to DC for a single peice of useful information is absurd.

Hear hear. Or here, here. Whichever. Not sure. Now we just need to establish the base DCs to begin with...

[ThiagoMartell]

I really don't think we need rules for that.

[whibla]

Part of the problem with knowledge skills is that they are intended to cover everything. This necessitates a very broad brush approach, which gives rise to some issues, notably the fact that details are lost and no account is given to regional or cultural variations in 'common' knowledge.
In addition, problems arise in the very definition of knowledge. Knowing something is a dragon, for example, isn't the same as knowing that it breathes a line of acid, and can cast spells as a sorceror of 15th level. In this respect, basing a character's knowledge on the hit dice of the creature they're trying to identify does have a certain sense or logic to it.
Finally, the rules do not include the possibility of false knowledge. As it stands, you either know something, or you do not, whereas, in reality, we all 'know' things that simply are not true. These false facts should also play a part in a complete treatment of, general or common, knowledge.

I realise from the title of the thread that this is an attempt to clarify and improve upon one specific aspect of the knowledge skill (albeit one that actually includes elements of at least 4 or 5 different knowledge skills), but, if you'll excuse me, I'd like to broaden this slightly. Laying a good foundation is critical to building a sturdy structure. With that in mind, it might be useful to follow an example from real life, just for a moment, in how we do actually structure knowledge. Most of us have probably heard of (or are at least familiar with the principles of) the Linnean classification system. Many spells (for example Polymorph Any Object) already explicitly refer to such a system, and I would suggest that this can form the basis for our simple, or common, knowledge checks, as follows:

Check DC's
DC 5*: Identify the kingdom (Animal, Vegetable, Mineral) of a creature or object.
DC 10: Identify the class (Mammal, Fish, Insect, Reptile, Fern, Fungi, Metal, etc.) of the creature or object.
DC 15: Identify the type (Horse, Cow, Human, Dwarf, Dragon, Oak Tree, Bamboo, Silver, etc.) of the creature or object.
DC 20: Identify the sub-type (Arabian, Friesian, Dark Elf, Fire Ant, Red Sandstone, etc.) of the creature or object.
DC 25: Checks at this level can never be considered common knowledge.

Modifiers
Racial / Cultural knowledge: +5
Regional knowledge: +5
Synergy bonus: +2 (per synergy)
Distracted: -2

Check Results
DC: Can name the (type of) creature or object.
DC+5: Can relate (some) general information regarding the (type of) creature or object.
DC+10: Can relate (some) specific information regarding the (type of) creature or object.
Each additional +5 will net further infomation. Note, the information gained is additive to the base classification DC's, and cumulative.

Examples
1. A multicultural human sees a Dwarf, takes a 10 (+5 cultural, +5 regional) and knows not only that it is a Dwarf, but that Dwarves can see in the dark, and have some affinity with stone. The specifics of this affinity, or the range of darkvision are not known.
2. An elf sees a dryad in the forest, takes a 10 (+5 cultural, +5 regional) and knows that it is a fey creature (DC 10), that it is called a Dryad (DC 15), and that it is associated with a particular tree (DC 20) and that it can use magic (DC 20). Additional details would require further study, or direct experience.
3. A farmer sees a large creature grab one of his sheep, and fly off. Taking a 10 he identifies it as reptilian (DC 10), and he runs back to town and tells everyone that a dragon raided his field. The adventuring party sent to kill the beast is disappointed to find the Wyvern (DC 15 to name) has no treasure.
4. A desert wanderer is ambushed by a giant scorpian. He knows the scorpian (DC 15 to name) is an insect (DC 10) that has a hard exoskeleton (DC 10+5), and a poisonous tail (DC 15+5). The nature of the poison is not known, but it's likely to be unpleasant.

I'm sure there's room for improvement on the above, and obviously the base numbers themselves are open for debate, as is the nature of common facts vs. specific facts (I'd tend to think in terms of a dragon (DC 15) has a breath weapon** (DC 15+5), and some immunities (DC 15+5), while an old black dragon (DC 20) has all the aforementioned, e.g. is immune to paralysis (DC 15+10), but it's breath is a line of acid (DC 20+5), 100' long (DC 20+10), and casts as a 7th level sorceror (DC 20+10)).
I look forward to seeing what you think, any alternate suggestions, or modifications to improve things.

*A creature of animal intelligence (-4 penalty), taking a 10, can recognise the difference between food (vegetable) and a rock (mineral), unless it's distracted (-2 penalty). This tallies with my observations of my, admittedly stupid, dog.
**All dragons breath fire, right?

[Roguenewb]

Frankly, for basic monsters, I just let a player know what the hell it is. Then, I answer questions based on the total magnitude of the check and how I feel about rarityu. AKA I just bull**** this relatively small part of the game.

[TuggyNE]

Check DC's
DC 5*: Identify the kingdom (Animal, Vegetable, Mineral) of a creature or object.
DC 10: Identify the class (Mammal, Fish, Insect, Reptile, Fern, Fungi, Metal, etc.) of the creature or object.
DC 15: Identify the type (Horse, Cow, Human, Dwarf, Dragon, Oak Tree, Bamboo, Silver, etc.) of the creature or object.
DC 20: Identify the sub-type (Arabian, Friesian, Dark Elf, Fire Ant, Red Sandstone, etc.) of the creature or object.
DC 25: Checks at this level can never be considered common knowledge.

While your basic idea seems reasonable enough, and interesting, requiring a DC 15 check to know that a stubby little humanoid with an axe is a Dwarf is at least 5 points too high, and requiring DC 20 to determine the breed of horse is also a bit much. (I would have a moderate chance IRL of distinguishing Belgian from Arabian from Quarterhorse, and I have no more than one cross-class rank in the skill, +2 int, and no Skill Focus. So DC 18 is the highest it should go.)

Modifiers
Racial / Cultural knowledge: +5
Regional knowledge: +5

+5 modifiers are terrifyingly high, in my opinion, being larger than the difference between, say, using the best possible locksmithing tools and improvising the crudest tools from what's at hand.

The rest of the proposed system could use some more tweaking along the same lines, but is probably at least usable.

That brings up another problem. By rule you get the ID at some DC and "a bit of useful information about that monster. For every 5 points by which your check result exceeds the DC, you recall another piece of useful information."

Yeah, I'm going to edit in another entry in the list to get rid of that particular really stupid rule. In fact, I think I'm going to operate on the principle of every +2 revealing more.

I think perhaps the biggest problem with knowledge checks to identify creatures is the untrained rule that prohibits you from succeeding on a knowledge check of DC 11 or higher if you're untrained in the skill.

An excellent point. As I see it, consistency with the other trained-only skills is important, and the four that have DC 10 tasks (Knowledge, Handle Animal, Disable Device, Sleight of Hand) all allow untrained users to attempt those and only those. So rather than allowing up to DC 15, say, I'd prefer to adjust DCs down to allow at least basic identification ("this is a dwarf") untrained below 10.

Alternatively, just outright remove the trained-only status, if that's justified, although that's a pretty big change.

As for things like dragons, it seems reasonable to me to handle the scaling thusly: if you can identify the youngest age category, you know the basic abilities, but you have to identify an older age category if you want to know things that the dragon can't do until those age categories. Say you're fighting a great wyrm blue dragon, but you only pass the knowledge check to identify a young one. You know basics blue dragon stuff, but you don't know about it's spell casting, its DR, its SR, or most of its spell-like abilities.

I'm torn here. There's certainly some sense to higher-HD creatures being more exotic and hence harder to identify. Honestly, it's only Dragons for which it's usually a problem. Maybe a caveat that for Dragons you use the smallest form, and for everything else you use pre-Advancement HD?

I actually intended to include this in the guidelines, that closely related creatures (the various listings that make up the Red Dragons, for example) are identified primarily based on the lowest one. But apparently it wasn't phrased well enough.

I agree, I'd love to simplify it somehow. Honestly, the easiest way would have been if the designers included a line in each statblock on the DC to identify. As they have not, I see few alternatives, otherwise, as mentioned above, icewind dale barbarians can identify camels. But if someone else comes up with a creative solution I'd be willing to entertain it.

The monster lore project I linked to included a bunch of these, in the MMIV style; unfortunately, the DCs were still wrong in many cases: a mule is specifically listed as DC 11, it's DC 30 to determine that the elf you see has relatives like half-elves and drow, and so on.

Frankly, for basic monsters, I just let a player know what the hell it is. Then, I answer questions based on the total magnitude of the check and how I feel about rarityu. AKA I just bull**** this relatively small part of the game.

I'd prefer that the rules help the DM a good bit more, rather than requiring analysis of all the intersecting mathematical, fantasy cultural, fantasy biological, and balance considerations. And as Rich Burlew said about Diplomacy:

The "patch" for the last two complaints is often the belief that the DM should apply circumstance penalties as he sees fit. My problem with this is without any guide as to what those penalties should be, it basically boils down to the DM thinking, "Do I want to give them such a huge penalty that they can't succeed, or not?" But I rarely have a preference. I don't decide whether I want someone to be persuadable, I want a rule system that lets me determine it randomly. It makes it very difficult to "wing" an adventure when there is no system for determining how to assess modifiers to this skill. Is that circumstance worth a -1? A -4? A -15? There's no guidelines given. In short, I want tools to use in the game, not a blank check to do what I want. I can already do what I want.

---

1
Another thread brought up the distinction between basic and combat-usable Knowledge checks, which is not made in the actual skill description, but is conceptually important. In other words, anyone should have a chance to identify all the creatures that live nearby, but the ability to remember specific abilities that are relevant in combat is not necessarily quite so easy. (Still: red dragons breathe fire, and that should not be especially difficult to determine; DC 15 at the very most.)
The main thing to watch out for here is metagaming potential: anyone who hears "dwarf" can remember the metagame information rather well, though their characters don't necessarily have the information.

Pulled in from RACSD:

Spoiler

Show

Elementals probably should have frightful presence, or a similar ability. They're aspects of primordial chaos, things which are alive that are made of stuff that shouldn't be alive. The more elder and powerful they are, the more nearly they resemble the hoary and eldritch lords of their planes - indeed, if you know just a little about them, they're even more terrifying, because they're literally all small pieces of those ultra-elemental horrors. The only name I remember is Cryonax and there aren't always cold elementals, but if we assume there are, each and every cold elemental is really just a tiny piece of Cryonax, serving as his eyes and ears throughout the world he intends to conquer. So yeah, they ought to be frightening, and more so the more uber they are.

So... how many creatures get frightful presence, anyway? Animated objects? Arrowhawks? The various grades of Formian? Salamanders? Sharks? Snakes? Xorns? There's a lot of scaling weirdness going on.

This is part of why there's a cap on CR scaling. If you are able to think of an easy way to kill them en masse, they aren't that scary. It's when that knowledge check fails that the fear sets in, and if you've invested lots of ranks in study, then you can't fail; you're a veteran demonologist and you know that dretches are pathetic, so there's no chance of you being frightened by the idea of them, even though there's some chance of you being killed if they get the drop on you.

Well, I was mostly referring to a case where it's actually a challenging at-level fight, and while Knowledge can help stem the tide, it's not automatic. Say a level 7 party being swarmed by 8 dretches. In that case, it's reasonable for a Knowledge check to reveal that use of most elemental damage will be less effective, hence the sound burst.

(stuff on phobias)

This is a pretty solid attempt at making sense of the confused muddle that I consider the existing rules to be. But the argument relies on the essentially arbitrary nature of fear, and a heavy (inferred) connection between that and ability to remember and apply knowledge; given that, why is this supposedly arbitrary "fear penalty" so smoothly scaling with actual danger, as measured by the admittedly terrible scale of HD or CR? In other words, a scholar conveniently is (universally!) more petrified by a creature that is actually more dangerous; this reaction is not especially reliant on subjective notions, and applies evenly whether a monster is subtle (medusa), aggressive (hill giant), completely baffling (phasm), actively using fear (sea hag), or outright trying to calm you down.

And, of course, it still falls foul of immunity to mind-affecting, immunity to fear, etc; it's irrational for a construct, even a living construct, to be so irregularly affected by the presence of an enemy, but warforged would certainly not get an exception. Same with undead making their checks. And what happens when scrying on something from a distance?

As I see it, it's trading a terrible explanation (HD = rarity of knowledge) for a slightly less terrible (HD = how much it shakes you), with some bizarre rules interactions on top.

[whibla]

While your basic idea seems reasonable enough, and interesting, requiring a DC 15 check to know that a stubby little humanoid with an axe is a Dwarf is at least 5 points too high, and requiring DC 20 to determine the breed of horse is also a bit much. (I would have a moderate chance IRL of distinguishing Belgian from Arabian from Quarterhorse, and I have no more than one cross-class rank in the skill, +2 int, and no Skill Focus. So DC 18 is the highest it should go.)

We live today in a knowledge focused world, where information is simply a click away, and turning on the science channel exposes us to things our medieval ancestors would have either had no clue about, or heard about only as myth or rumour. We spend the majority of the first 18+ years of our life learning knowledge based skills. I have no idea as to your background, but from mine, despite living in farming country for a couple of decades I wouldn't stand a chance of reliably and correctly identifying most breeds of horses, cows, or even dogs. Now, I did say in the original post that the numbers were open to change, but the question is by how much? Is a Dwarf (correctly identified by race) really that recognisable, to someone who's never seen one before? We, as long time players of D&D have considerably more than 1 cross-class rank in Knowledge (Fantasy), as well as cultural bonuses (LotR etc.). Can a semi-literate peasant, who's spent his life working his fields in a human only domain really, definitively, identify a Dwarf on sight, by taking a 10? Food for thought...

+5 modifiers are terrifyingly high, in my opinion, being larger than the difference between, say, using the best possible locksmithing tools and improvising the crudest tools from what's at hand.

It would seem so at first glance, yes, yet it's less than the natural variability between people at opposite ends of the IQ (for want of a better word) scale. Also, using the example you give, the difference between using a couple of paperclips, or a credit card, and some exceptionally engineered 'picks', or a metal card, is almost as much about the durability of your tools as it is their usefulness. Yes, they make the job easier, but at the end of the day a spike of metal is a spike of metal. I'm not sure it's helpful to consider modifiers to mean the same thing (or the same quantity of thing) when comparing widely different skills. Again, just a thought...

Yeah, I'm going to edit in another entry in the list to get rid of that particular really stupid rule. In fact, I think I'm going to operate on the principle of every +2 revealing more.
...
Alternatively, just outright remove the trained-only status, if that's justified, although that's a pretty big change.

I tend to agree that a single piece of knowledge per 5 points is ridiculous. That was one thing I was trying to mitigate by allowing the same roll to expose detailed knowledge (of dragons for example) and common knowledge (of red dragons for example) at the same time. Even so, additional information per +2 is probably a reasonable rule.
With regards removing the trained-only nature of knowledge skills, I actually think most, if not all, should remain trained only, the only exception being perhaps Knowledge (Local). I would prefer to see the (regional etc.) modifiers I gave as a negative modifier to the DC of the check, rather than raise the DC that untrained people can obtain, or allow anyone to be able to read a set of arcane runes (for example) simply because they happened to roll well.

Finally, I would just like to emphasise: perspective. The rules have to apply equally to everyone, from their perspective, not ours, with our overview of the game system, and history as game players. Ask not only "If I were a ill educated dirt farmer, who has never travelled more than 12 miles from home in my entire life would this make sense?" but also "If I were a 10th level druid with 10 ranks in knowledge (nature) would this make sense?" etc.

[Urpriest]

I really don't think we need rules for that.

This.

Actually, here's a decent rule: your character knows, by default, what any given type of monster looks like, as well as any traits that are reasonably common in folklore. I could see checks being made to determine specific resistances, SLAs, and the like, but knowing what a creature is, unless it's disguised as something else, should be automatic.

I'm not proposing this from a realism standpoint. Realistically, there can never be enough knowledge skills to cover the variety of possible adventurer backgrounds. Instead, I'm proposing it from a gamist standpoint. D&D 3.5 is a game of fast-paced, deadly combat. There isn't enough time for characters who don't know what they're facing to poke the monster and stall until they figure out how to fight it. D&D is chess played with immunities and BFC, and chess is the sort of game that it's simply unfair to play with imperfect information.

[TuggyNE]

We live today in a knowledge focused world, where information is simply a click away, and turning on the science channel exposes us to things our medieval ancestors would have either had no clue about, or heard about only as myth or rumour. We spend the majority of the first 18+ years of our life learning knowledge based skills. I have no idea as to your background, but from mine, despite living in farming country for a couple of decades I wouldn't stand a chance of reliably and correctly identifying most breeds of horses, cows, or even dogs. Now, I did say in the original post that the numbers were open to change, but the question is by how much? Is a Dwarf (correctly identified by race) really that recognisable, to someone who's never seen one before? We, as long time players of D&D have considerably more than 1 cross-class rank in Knowledge (Fantasy), as well as cultural bonuses (LotR etc.). Can a semi-literate peasant, who's spent his life working his fields in a human only domain really, definitively, identify a Dwarf on sight, by taking a 10? Food for thought...

Indeed. Especially I'm beginning to wonder why it's possible to take 10 on Knowledge checks.

It would seem so at first glance, yes, yet it's less than the natural variability between people at opposite ends of the IQ (for want of a better word) scale. Also, using the example you give, the difference between using a couple of paperclips, or a credit card, and some exceptionally engineered 'picks', or a metal card, is almost as much about the durability of your tools as it is their usefulness. Yes, they make the job easier, but at the end of the day a spike of metal is a spike of metal. I'm not sure it's helpful to consider modifiers to mean the same thing (or the same quantity of thing) when comparing widely different skills. Again, just a thought...

Your examples gave not one but two +5 modifiers, and then added zero or more +2 modifiers on top. That just feels far too swingy. (In the absence of a complete system, all I can really give is generalities.)

With regards removing the trained-only nature of knowledge skills, I actually think most, if not all, should remain trained only, the only exception being perhaps Knowledge (Local). I would prefer to see the (regional etc.) modifiers I gave as a negative modifier to the DC of the check, rather than raise the DC that untrained people can obtain, or allow anyone to be able to read a set of arcane runes (for example) simply because they happened to roll well.

Technically, pretty sure reading runes would be Decipher Script, not Knowledge: Arcana, but I'm sure there's another similar example.
The main objection I'd have to negative modifiers to DCs based on circumstances is that they are non-standard and therefore confusing; the reason for their inclusion is not immediately obvious, so someone might carelessly standardize them and disrupt the intended design.

A specific variant would be to allow any identification check to be made untrained; that would at least be more understandable, if still a bit non-standard.

Finally, I would just like to emphasise: perspective. The rules have to apply equally to everyone, from their perspective, not ours, with our overview of the game system, and history as game players. Ask not only "If I were a ill educated dirt farmer, who has never travelled more than 12 miles from home in my entire life would this make sense?" but also "If I were a 10th level druid with 10 ranks in knowledge (nature) would this make sense?" etc.

This is true, and I tried to figure out a way to balance that into my criteria, but I don't think it came through very well.

Actually, here's a decent rule: your character knows, by default, what any given type of monster looks like, as well as any traits that are reasonably common in folklore. I could see checks being made to determine specific resistances, SLAs, and the like, but knowing what a creature is, unless it's disguised as something else, should be automatic.

I'm not proposing this from a realism standpoint. Realistically, there can never be enough knowledge skills to cover the variety of possible adventurer backgrounds. Instead, I'm proposing it from a gamist standpoint. D&D 3.5 is a game of fast-paced, deadly combat. There isn't enough time for characters who don't know what they're facing to poke the monster and stall until they figure out how to fight it. D&D is chess played with immunities and BFC, and chess is the sort of game that it's simply unfair to play with imperfect information.

I'm not entirely sure how to react to this; on the one hand, I have a strong urge to make the system "correct"; on the other, I recognize the necessity of keeping it simple enough to use; and on yet a third hand, I dislike metagaming, which either an extremely simplified or an extremely complex system is likely to encourage (for different reasons, of course).

Well, I guess I'll keep thinking about this. I may revise my suggested identification scheme after a bit.

[Ashtagon]

DC 15: Identify the type (Horse, Cow, Human, Dwarf, Dragon, Oak Tree, Bamboo, Silver, etc.) of the creature or object.

Farmer Giles can't tell a horse from a cow. Or, indeed, from his wife.

[Andorax]

It's been touched on to some degree here, but I'll toss out something I posted over in the "Common Sense" rules thread:

If you can make a successful knowledge check (PHB 78) to identify a fundamentally similar creature's traits, you can successfully identify all of the common traits between that creature and the one you are observing.

The exact definition of fundamentally similar may vary by DM, but at its most basic level, it includes all versions of a creature that advances by ages or age categories (for example, the True Dragons MM68-88, Neogi Spawn -> Adult Neogi MM2 159), all creatures who are called the same thing with only a size category distinction (for example, the Elementals MM 95-101), and any creature that is described as being a "Lesser" or "Greater" version of another (Stone Golem -> Greater Stone Golem, Fihyr -> Great Fihyr (MM2 100)).

Now this doesn't answer (or even attempt to answer) how to fix the DCs, or what sort of checks you'll need to determine specific traits...other good ideas have been, and likely will be, posted to get at that issue.

This is just an attempt to simply codify picking up on related creature knowledge.

[mattie_p]

How about something like this,

"For a knowledge check to identify a creature, the following special rules apply.

* The DC for a creature's own type and the animal type is DC 5+HD, all others remain unchanged.

* Even untrained, a character an take 10 to identify a creature, as per the normal rules of taking 10.

* As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check."

In other words, you can't take 10 in combat. Joe commoner will always be able to identify his wife and his mule (and tell them apart), but might not be able to determine whether it is an ogre or a fire giant who is raiding his cattle.

After the fact, he goes to the local bar and starts describing this giant. 10 of his buddies sit around, drinking ale and scratching their head, until they come to a collective decision.

Sound remotely interesting to anyone?

[whibla]

* As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check."

In other words...Joe commoner will always be able to identify his wife and his mule (and tell them apart), but might not be able to determine whether it is an ogre or a fire giant who is raiding his cattle.

After the fact, he goes to the local bar and starts describing this giant. 10 of his buddies sit around, drinking ale and scratching their head, until they come to a collective decision.

The D&D version of the Wisdom of Crowds. I like it! On balance, I think I prefer this version to mine: Simpler, elegant, and with some real world logic behind it. Good call...

[Mnemnosyne]

How about something like this,

"For a knowledge check to identify a creature, the following special rules apply.

* The DC for a creature's own type and the animal type is DC 5+HD, all others remain unchanged.

* Even untrained, a character can take 10 to identify a creature, as per the normal rules of taking 10.

* As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check.

* When aided by others, knowledge checks to identify creatures can be made untrained even if the DC is higher than 10."

Good idea, but needed one more exception in order to make it work, because you didn't allow for knowledge checks to work for DC 11 or higher regardless of the Aid Another. So in your original example, Joe and his 10 friends would get a Knowledge check of 30, but still be unable to identify the ogre/fire giant because of the untrained rule about DC's higher than 10.

[MukkTB]

D&D is chess played with immunities and BFC, and chess is the sort of game that it's simply unfair to play with imperfect information.

In chess you don't wonder if there's a queen hiding around the corner. In chess you don't have to roll to see if your piece can take another piece. In chess your ultimate fate isn't dependent on the continued benevolence of another player pulling all the strings.

D&D 3.5 is much more simulationist than chess, and much less gamist.

[mattie_p]

Good idea, but needed one more exception in order to make it work, because you didn't allow for knowledge checks to work for DC 11 or higher regardless of the Aid Another. So in your original example, Joe and his 10 friends would get a Knowledge check of 30, but still be unable to identify the ogre/fire giant because of the untrained rule about DC's higher than 10.

Good call. I like it. That was implied in my suggestion, but we really do need to specify it.

Any other criticism or suggestions? I think we're on the right track here, and I think that the "aid another" should get a +1 circumstance bonus per pint of ale consumed!

No, not really. Then you know the fire giant's mother's maiden name, eventually.

[mattie_p]

Oh, two more things.

1) As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes one hour of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check.

This reflects people searching their memory from when they were a hired sword on the merchant caravan from the city of Sandstone to the city of Stormsea and saw something. Also, this reflects the time that people have to spend sitting around talking about whatever to remember a key fact ("Yeah, I left my keys on the coffee table!")

The second is optional, but I am totally going to implement it.

2) On an untrained knowledge check, where no one has any specific knowledge in the pertinent field, the group consensus as a 5% chance of being almost completely wrong. Roll a d20 in secret, on a natural 1, the group is able to identify the type but is completely wrong in its identification of the creature. None of the group can be later convinced that they were wrong, despite incontrovertible evidence.

This reflects that group think can sometimes come to a completely incorrect conclusion based on the facts. "So you say that the dragon was red. That means the dragon is susceptible to fire damage. Dude, it's true! You thought you saw the dragon breath fire, but everyone knows that cold burns as well."

Feel free to change the percentage if you wish. I call this the Colbert effect.

[KillianHawkeye]

It's probably best to work upward from the simplest DCs to the more complex ones, because it's easy to guess that most people should know what a mule is (so the DC has to be no more than 10), and so on; then, using comparisons, work up to the big stuff.

I just want to say that while I know WHAT a mule is, identifying it as such when encountered in the wild (and subsequently knowing anything useful about its habits or capabilities) is not going to be so easy for me.

[Mnemnosyne]

I just want to say that while I know WHAT a mule is, identifying it as such when encountered in the wild (and subsequently knowing anything useful about its habits or capabilities) is not going to be so easy for me.

You, however, do not live in a subsistence-based pre-industrial agrarian society. A more apt comparison in our modern lives would be to be unaware of the capabilities of a pickup truck. I don't own a pickup truck, and you may not own one either, but I'm sure we've both seen many pickup trucks and have a pretty good idea of their capabilities. So it would be with your average commoner in a D&D world. He may not own a mule, but he has certainly seen a great many of them and has a reasonable understanding of how they behave and what they can do.

[137beth]

9+CR for most types and 4+CR for humanoids, animals, and giants

Dragons should also probably be lower. Almost everyone knows what a dragon is, even if its CR is high.

[TuggyNE]

How about something like this,

"For a knowledge check to identify a creature, the following special rules apply.

* The DC for a creature's own type and the animal type is DC 5+HD, all others remain unchanged.

* Even untrained, a character an take 10 to identify a creature, as per the normal rules of taking 10.

* As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check."

Also, humanoids should always be 5+(scale), because whatever type you are, humanoids are common and easy to identify.

1) As an exception to the rule where no character can take 10 in an attempt to aid another, a character can use 2 minutes one hour of time to take 10 on the check to aid another. There is a limit of 10 additional people who can aid on this type of knowledge check.

This reflects people searching their memory from when they were a hired sword on the merchant caravan from the city of Sandstone to the city of Stormsea and saw something. Also, this reflects the time that people have to spend sitting around talking about whatever to remember a key fact ("Yeah, I left my keys on the coffee table!")

This ... probably makes sense, although I think it could use a bit more tweaking.

A more apt comparison in our modern lives would be to be unaware of the capabilities of a pickup truck. I don't own a pickup truck, and you may not own one either, but I'm sure we've both seen many pickup trucks and have a pretty good idea of their capabilities. So it would be with your average commoner in a D&D world. He may not own a mule, but he has certainly seen a great many of them and has a reasonable understanding of how they behave and what they can do.

Indeed. One thing to consider is that, to my understanding, D&D seems to portray a more highly educated peasantry than historically existed, so while they would obviously have a good familiarity with the various beasts of burden and such, some of them would apparently have an unusual amount of knowledge about more exotic creatures, if only fantastic predators like ankhegs. (Commoners seem to be considered literate by default, for one thing.)

Dragons should also probably be lower. Almost everyone knows what a dragon is, even if its CR is high.

I'm a bit divided on this. True dragons are mostly well-known, so should probably receive an adjustment; lesser dragons not so much, though. Also, anyone houseruling in fake information should include extra for true dragons, because their notoriety breeds rumors both true and false.

---

1
Switching gears for a bit: what mechanism best represents the increasing rarity of reliable information on rarer or perhaps more dangerous creatures? Are more dangerous creatures inherently less likely to spread accurate information around, and if so, what's the scaling rate? (In rules terms: HD, CR, or something else?)

I think it's fair to say that knowledge of an ancient green dragon's ability to dominate should be more difficult to acquire than the realization that a mule may kick you hard. But is a DC 28 (10 - 5 + CR 21 + 2) check vs DC 8 (5 + CR 1 + 2) check really correct? Are only dragon specialists likely to be aware of this?

[georgie_leech]

Perhaps a misinformation check of some sort? Say, a percentage-based check modified by the CR and creature type in question; things like outsiders, dragons, undead and other "famous" powerful creatures likely have many false stories and legends about them (I hear tell that vampires actually sparkle in daylight, which makes them embarrassed!"). The chance could be reduced by either your knowledge check or skill points (x4 or some other modifier), to avoid the case of a so-called expert being dreadfully misinformed on the subject of his expertise.

[whibla]

Also, humanoids should always be 5+(scale), because whatever type you are, humanoids are common and easy to identify.

In truth I'm not sure I agree with this. Mattie_p's idea of DC 5+ for creatures of the same type covers this for a large proportion of pc races, but otherwise I might dispute that the difference between a halfling and a gnome, or a lizard man and a saurial, is that obvious to a non-humanoid. The reason for lowering the DC is purely to make untrained identification possible, surely, not to weaken the skill?

I think it's fair to say that knowledge of an ancient green dragon's ability to dominate should be more difficult to acquire than the realization that a mule may kick you hard. But is a DC 28 (10 - 5 + CR 21 + 2) check vs DC 8 (5 + CR 1 + 2) check really correct? Are only dragon specialists likely to be aware of this?

I'm not sure I'd agree, either, that any other legendary monsters, even dragons, should have the DC to identify lowered. Misidentification, or misdirection by the DM, is an inherent part of the skill. "You see a large flying lizard" could mean it's a dragon, or it's a wyvern, or it might even be a pterodactyl. Untrained (seen no life-like pictures in books) I do not think the distinction is obvious, and the colourful language of a bard in relating the saga of Siggurd and the Great Wyrm could be as much a hinderance as a help.

With regards the specifics of DC's relating to information, rather than identification, you can reach a DC of 28 at 3rd or 4th level with a bit of luck, or access to a library (enabling you to take a 20). The -5 to the check you've included is uneccessary, and lessens any reason to take more than a single point or two in the knowledge skill(s). After all, if your party does suddenly encounter an ancient green dragon, with no forewarning, and no pre-planning, instant awareness that it might dominate one of them is likely to be the least of their worries. Having said that, CR as opposed to HD might be a sensible compromise for information scaling. That is a good idea.