I got into a big argument on another forum about the fallibility of knowledge checks. I was under the impression that they represented things a character had learned by virtue of living life in the campaign setting, so that he could act on that knowledge whether or not his player had that knowledge. This meant that things you learned were relative to your character.

Other people seemed to be of the opinion that if you invested resources in a knowledge skill you would be rewarded with perfect information whether or not there was any explanation of how you obtained it.

In games that use a "knowledge check" of some kind to determine what a PC might know about something, what is the default expectation of the nature of that knowledge - subjective or objective?

I'm using the system agnostic tag because I'm not interested in a "system specific answer" just to the system I'm using, I'm interested in hearing about this across a variety of systems. Is it something that varies from system to system? Is it a group preference thing?

  • 9
    \$\begingroup\$ This is very system-dependent. Consider retagging to fit your actual problem scope, even if you think this could be more general. There are too many system nuances to have a general answer to this. \$\endgroup\$ – SevenSidedDie Jan 12 '13 at 4:41
  • 10
    \$\begingroup\$ I agree. In fact, so long as it remains system-agnostic, I'm voting to close it as Not A Real Question. It's impossible to even begin to answer this in general. It's a matter of the rules and expectations of the system, the setting, and the group playing them. \$\endgroup\$ – KRyan Jan 12 '13 at 5:31
  • 1
    \$\begingroup\$ I think this works fine as system-agnostic, because this is PRIMARILY the expectations and play style of the group, not the rules themselves. \$\endgroup\$ – mxyzplk - SE stop being evil Jan 12 '13 at 15:07
  • 3
    \$\begingroup\$ @mxyzplk Only in some systems. Other systems do have rules answers for this, eg. Burning Wheel, Dungeon World. This Q reads to me like a D&D-dependent question. \$\endgroup\$ – SevenSidedDie Jan 12 '13 at 16:44
  • \$\begingroup\$ Just because there's a couple indie games that handle it explicitly, the 95% bulk of all the trad games out there all handle it the same way. Those that work differently might provide interesting insight to an answer, not invalidate the possibility of one. \$\endgroup\$ – mxyzplk - SE stop being evil Jan 12 '13 at 17:51


I do not think this question makes a lot of sense as a question, and I am answering on the basis of systems with which I am most familiar. I strongly suspect that there are other systems where what I say would be flat-out and explicitly wrong. However, there have been claims that 95% of systems handle these things the same way, which, if true, may make this answer useful.


A successful check should never have the character positive of something that is false. A mediocre check (or a very high difficulty) may result in uncertainty, but the character should be aware that he is uncertain unless he rolls badly. Even then, rolling badly is intended to mean more “you don’t know” than it is “there’s no such thing as a jabberwock; everyone knows that!

On the other hand, being sure of oneself, when one is in fact wrong, is pretty common, in stories and in real life. As a result, it may be an appropriate thing to have in a game. Again, it is much more common when the person in question knows a little bit – the Dunning-Kruger effect, for example – so this should usually be the realm of low rolls. And in many games, it is inappropriate to punish characters for doing well, but not well enough, even though this is a common theme in reality (“a little knowledge is dangerous”). Being misled ought to be a function of significant failure.

If you have a secret the players are not meant to know, that’s a plot point and really shouldn’t be a matter of rolling. Rolling dice when the answer is predetermined just slows down play.

Regardless of how you set it up, however, it is important to explain it to your players ahead of time, ideally before the game starts. Obviously, it is not a thing to bring up just as you’re deceiving them, since that would be a dead give-away, but a lot of players will assume that by the nature of the game and its rules, Knowledge checks are infallible. You need to explain that what they are told upon rolling Knowledge is what they remember, believe to be true, think they know, etc.

If certainty-in-the-inaccurate is to occur only on very poor rolls (which I recommend), you should also probably be rolling the dice in secret for Knowledge checks. In which case you also have to describe how certain the character is: “you’re absolutely certain” for high rolls, “you’re pretty sure it went like this” or “common knowledge says that this is so” for medium rolls, “you just don’t know” for low rolls, and “you’re absolutely certain (and you’re wrong)” for the lowest rolls, or perhaps only on the lowest roll.

  • \$\begingroup\$ +1 for the last paragraph; but you need to distinguish between What do I know about this? which should probably be rolled in secret to allow for the possibility of misinformation as well as/instead of the real facts, and Do I know...[the name of the most powerful wizard in town] which works better as an open check made by the player. I think this choice depends more on the group than the system. \$\endgroup\$ – Tim Lymington Jan 12 '13 at 23:12
  • \$\begingroup\$ I like how you inserted the Dunning-Kruger effect. ^.^ One should also note that you might roll very low on a knowledge check about a common Ogre and then roll very high on a check about an Aboleth or some crazy monster and that creates some semi-unrealistic situations. \$\endgroup\$ – Simanos Jan 5 '14 at 14:20

It Depends on the System

But for the most part the 'fallibility' of a Knowledge check is a matter of degree rather than outright misinformation. You can see this system in New World of Darkness, D&D 3.5, and Legend - a Knowledge (or Occult, or Academics, or...) check tells you information on a sliding scale of success, usually starting with something like "This is a dragon," and ending with, "This is a young adult black dragon, these are its weaknesses, it's immune to those things, and these are its supernatural powers." At no point do most systems tell you to give bad information to your players, excepting a few cases where the very low end of the scale begins with bad information that then becomes good information. It's usually considered extraordinarily poor form to give bad information to PCs if they make a successful check that is not some kind of dramatic failure, both because it shoots their fun in the foot and because it tells them that they wasted their valuable resource investment.

Additionally, most systems put Knowledge-style checks as a result of a combination of book learning, experience, training, research ability and access to reading material, and as a result a character might know things about phenomena or creatures they've never experienced directly. If it sounds a little strange, just think of it this way: lots of people have never seen a polar bear, but I bet you a whole lot of them can tell you about one.


It depends very much on system and group.

FATE allows you to use a skill check to make a declaration about the world that is relevant to that skill. So a Monsters skill would let you give the Werewolf a weakness to wolfsbane instead of silver, if you succeed on the check and the group doesn't think it's boring or silly. Whether you actually know that or find it out when you try to use it is a decision you as a player can make.

Misinformation can make for an interesting addition to a game. This depends on the group and the amount it's used. If everybody expects to know The Truth when they make their check then it will cause arguments if they find out that they believed A Lie. Using it outside a few limited circumstances that are important to the story will also detract from it.

To use misinformation well, you want to use it in situations where finding out the truth and changing your understanding is part of the drama. For instance - if the Werewolf isn't vulnerable to silver, there's got to be a reason why the player who made the Knowledge check thought it. Discovering what that reason is and following the consequences should be interesting. If it isn't, just tell your players the truth in the first place.


The Truth and...

Commonly in my groups, knowledge checks were more or less plot device sources. Due to the higher TN/DC that belonged to obscure information it is entirely easy to say "You once heard..." and the information that was correct from the check I would let them know they were solidly sure of, but there were rumors that nothing they'd experienced had confirmed. And that little bit on the end was always my way of throwing potential misinformation. Most of the uncertain caveats are rooted in superstition.

For example, in an L5R game a character knew that to some Dragon Clan monks had magical tattoos, and was able to recognize one and the power associated with it. However they had also heard an unfounded rumor that some tattoos were only taken by "fallen" monks.


Knowledge skills often represent a characters willingness to seek out information during his downtime.

A history buff will often be found in the library, talking to the locals and seeing the sights.

Some knowledges such as Nature, may be experience driven.

You only role play the exciting bits. There's another 14 waking hours each day in which the heroes do more mundane and personal things.

So if there's something to be known, select a difficulty and let them roll for it. If they succeed, they know what's relevant due to something they heard or read or saw, perhaps between scenes.

  • \$\begingroup\$ In fact, with a lot of questions regarding "Is it reasonable for a character to be able to do X?" my usual response is "He's a hero, so yes." \$\endgroup\$ – Hand-E-Food Jan 12 '13 at 5:47
  • \$\begingroup\$ I definitely agree that most skill points will come from stuff the character does automatically. I'm okay with assuming the wizard took a tome of arcane lore out of the library last time he was in town and had been reading it before bed. I'm not okay with him having access to Lord Evilguy's personal diary just because his knowledge: history is maxed out. What you suggest about this knowledge coming from offscreen experiences tells me the knowledge would be relative to the character instead of universal or omnipotent. \$\endgroup\$ – valadil Jan 13 '13 at 3:46
  • 1
    \$\begingroup\$ I get what you're saying. I wouldn't allow that. If the player insisted, I might allow an extremely difficult test to see if the character can deduce part of Lord Evilguy's plan, Sherlock Holmes style. At most, they might learn the next location they could find him or his minions. Intelligence is one of the hardest attributes to separate between the character and the player. An intelligent player playing a dumb-as-bricks barbarian is still intelligent, but I can't deny an acedemic wizard played by a fool from knowing things the fool would never consider. \$\endgroup\$ – Hand-E-Food Jan 13 '13 at 22:29

Speaking as a GM... yes. Sometimes. Really, "it depends" is the only truly accurate answer here, because there are a lot of different situations.

By way of an example, Lord_Gareth's example above talked about a knowledge check that reveals "This is a young adult black dragon, these are its weaknesses, it's immune to those things, and these are its supernatural powers."

That's fine and dandy, but what if it's a special dragon that lacks those weaknesses, but that isn't distinguishable from a normal one upon visual inspection? That kind of knowledge check is done when you see the thing, so you can't tell this is a fancy dragon nobody has ever heard of before from a normal one. In that case giving "true" information from the knowledge check requires some kind of divine insight: it's simply not possible for a character to know it on his own.

To me, the knowledge skills are things that you've learned somehow. What you know is not an infallible repository of perfect information that updates when I create a new spell that nobody on the planet has ever seen.


Not the answer you're looking for? Browse other questions tagged or ask your own question.