Monday, 9 January 2012

He that increaseth knowledge increaseth sorrow.

-C has been putting up a series of posts deconstructing skills in D&D. The most recent is about knowledge skills, and the problems attending them:

I am confused by the purpose of this skill [he writes]... I just fail to understand how [it] is supposed to be used in play. Here are the situations I can imagine.

There is a piece of information. It is either trivial and of no importance, or it is interesting providing some depth and background to the game and not vital, or it is a crucial piece of information.

In any possible conceivable case is the game improved by withholding any of the above information?

The answer to this question, for me at least, is a qualified "yes". There is one knowledge skill I can think of for which there is a justification, and one area in which it is concievable that the game can be improved by it. That's "Local Knowledge", or similar.

This much we know: there is a dilemma facing all DMs, especially the ones who create their own settings. On the one hand, you think your setting is interesting, and you think the players enjoy finding out about it. You also think it would really enrich the game if the players knew things that were going on, knew where things were, and knew who various NPCs were. But on the other hand, you know that your players' patience for reading or listening to infodumps about the setting are essentially nil.

This is where Knowledge skills come in: they allow you to give information to players, about the setting, that they want, at a time of their choosing. So we might imagine the following scenarios:

Bob has discovered an iron gauntlet of strange design in the dungeon. He doesn't know what it is. He rolls his "Local Knowledge" skill and the DM tells him he knows an armourer in a nearby town who could know more about it. The players go there and find out.

Gwendaline sees a strange monster she's never encountered before. After wisely fleeing, she rolls her "Local Knowledge" skill and the DM tells her of a sage who might know more. The players go to meet the sage and in exchange for performing some non-trivial task, get the information.

The party need to find an alternative route to such-and-such town because the regular road is blocked off by their enemies. Successful local knowledge rolls allow them to determine that there is a hunting trail through the forest. They go that way instead.

Hey presto! Local Knowledge makes the game better. It expands the players' knowledge of the setting in an organic way (because the information they glean is always something they actively want to know), and expands their options.

Now, of course, there's an obvious comeback that goes something like this: implicit in the above argument is the notion that for players to do or know anything, they have to roll to find out. But as -C says, it is hard to imagine how making players roll for this sort of thing really makes the game better. If the players fail their rolls, they don't get the knowledge, and the game suffers as a result, because the setting doesn't get fleshed out and the players' options don't get expanded. So why bother rolling? Why not just tell the players what they know if they ask? Or just have them ask an NPC?

There are a few reasons:

a) Having them just ask NPCs about stuff is time-consuming and quite boring. In the first scenario I included above, for instance, it would involve the players asking other NPCs they know if they know anybody who knows anything about gauntlets. Snore. Easier to roll and just say "yes" or "no", and if "no", then they can do the boring ask-other-people-if-they-know-anybody thing.
b) Having players ask the DM "What do I know about this?" or "Do I know somebody who does [x]?" or whatever puts a lot of narrative control in the DMs hands, and might prove an irresistable temptation to try to guide the "story" unless that DM is a paragon of objectivity. It might lead to scenarios in which the DM thinks to himself, "I'll tell the players this titbit of information/withold it because it'll make the story more interesting." That way madness lies. Having players roll for things like that keeps the DM honest.
c) Sometimes mystery is fun, and can make the game enjoyable in its own right. When you don't know what a monster can do, it makes it scarier. When you are carrying around some weird artefact that you don't know anything about, it makes it intriguing. When you don't know what lives in the hole in the mountain, you scout it out. Not everything ought to be mysterious, but there's nothing wrong with it in small doses.

I expect some people to disagree with this.


  1. This comment has been removed by the author.

  2. The thing is, I've played in the game that runs like that, and here's how it worked for me.

    "How do I raise my horse from the dead?"

    "You don't know."

    "Well how do I find out?"

    "Roll your knowledge: local."


    "The priests might know."

    I go to the priests.

    "Can you help me raise my horse?"

    "That's stupid! You can't raise a horse! Why are you bothering us?!"

  3. Oh, maybe I miss-communicated.

    A)I am not suggested eliminating the knowledge roll so they can talk to _NPC's_ about how to get their information. I'm suggesting they ask the DM directly instead.

    B) I don't see how the roll keeps the DM honest, because if they succeed, he's telling them the thing he would have told them without the roll?

    I am fine with C. I don't think it conflicts with knowledge skills or the lack thereof at all.

  4. Well, the first comment is just an example of dick DMing. No legislating for that!

    As for the rest...

    A) Okay, but then there's danger I talked about in B): isn't the temptation always there for the DM to then (maybe subconsciously) give out information in a manner he thinks is "interesting"? Like, "Oh, I'd like it if the players knew this, it would be fun, so I'll tell them"? And then isn't he just palette-shifting indirectly?

    That leads on to the dice roll thing. The dice keep the DM honest. If the PCs succeed, they know that information. If they fail they don't. That doesn't mean just telling them "You don't know and never will", of course. The players can then choose to try to find out the information through sleuthing or research or whatever.

  5. " I'd like it if the players knew this, it would be fun, so I'll tell them"?"

    God forbid the players should be pointed in the direction of fun... :P

    Is this some kind of extreme reaction to 4e's Tyranny of Fun? I mean, I'm running a pretty pure Gygaxian sandbox right now, but I certainly dump a whole load of rumours on the PCs that if followed up on are calculated to lead to fun - fun and/or TPK. >:)

  6. S'mon: I interpreted more as an attempt to maintain DM neutrality, and I mostly agree, because if the player succeeds on a roll then there seems an implicit contract that the information gained is accurate. These are good points in the argument 'for'.

    I have always wanted to like this skill (as Rolemaster's lore skills) but have never had a chance to play with it. It reminds me also of the Know roll in Call of Cthulhu - unfortunately, again, little actual play experience there).

  7. Had to read it twice...

    I think I get what you're saying about keeping the DM honest. BUT, the DM is the source of information in the game no? Sometimes players need a little direction, some background, or something that creates a little interest so that things don't degenerate into "roll dice, hack and slash..roll dice..."

    We use something similar to local knowledge in our games. Its a skill called "Mind for Miscellany". Only a couple of characters have the ability, it was taken from a list of random abilities we rolled for when creating our characters. It works in a similar way, we see some strange graffiti scrawled on the walls of a local temple, a player with the miscellany skill rolls to see if they "know" anything about it. If so, the DM can fill in the blanks. If he fails his roll, there is no instant gratification. Either the players log it for another time, or find a different approach.

    I think either of these techniques takes the heat off of the DM, or "keeps him honest". As long as it creates interest, or adds detail(fun) to the game, I'm for it.

  8. I've been reading -C's posts about skill reconstruction, and i have agreed mostly with the criticisms raised.

    About halfway through the series though I started GMing type 4 and the idea of trained skills seems to help both ways. The way I have been adjudicating them is that characters with trained in something get 'GM's Bonus' which is where I wi,l directly tell them the things you would normally require a knowledge roll for without a roll. Hence the party gets the critical information, but it always comes through the logical party member for that information (the wizard for arcana or history). This works in tandum with rolling for really esoteric knowledge or to abstract whether or not a player, say already knows a smith in this village, or forthe abstraction required for a stealth v perception check.

    This way of adjudicating skills at least for me, has worked really well in getting rid of the 'roll knowledge 'I get a 5' 'okay you don't know anything situation.

    Just my two gp.

  9. "b) Having players ask the DM "What do I know about this?" or "Do I know somebody who does [x]?" or whatever puts a lot of narrative control in the DMs hands, and might prove an irresistable temptation to try to guide the "story" unless that DM is a paragon of objectivity."

    This bit for me. I like having a random roll to see if the players get a real clue or a line of BS (or a mix of both), for the same reason that I like wandering monster rolls, it takes some things out of my hands and makes me more of a Referee than a Storyteller (which I like for D&D-style games). It's mostly psychological, but it makes the game feel more real if the players are getting screwed over by dice rolls than by me, even if I'm the one who chose what the dice rolls will do.

  10. I tend to use pass/fail knowledge rolls in the following cases:

    #1. Where success will give a mechanical advantage vs a Known Unknown. Eg solving a puzzle, disabling a trap, succeeding in a negotiation, defeating a monster.

    #2. An opportunity for me to establish previously undetermined campaign facts - ie Make Sh*t Up. If the player makes the Kno roll, only then do I decide what the facts are - say, the religious rites of the Church of Bane - and tell the player.

    I try to avoid using pass/fail Kno rolls for hooks that will lead to adventure opportunities and More Fun. I would try to avoid a Pass/Fail Streetwise check for rumours of exciting adventure in the Beggar King's Lair. I'll use the Streetwise check for info that would be useful in defeating the Beggar King - type #1 above, or for previously undetermined facts about the Beggar King - type #2 above.

  11. I think there's another piece to this discussion - player enjoyment.

    Personally - and this may not hold true of all players - I get more pleasure in using a skill/beating the dice then just being fed information by DM fiat.

    When you roll - successfully! - to gain knowledge that helps meet a challenge I feel self-satisfied -'I played that well'.

    On the other hand, if you just ask the DM neither getting the answer or not is particularly satisfying. If the DM says 'yes, you know x, y, z' then you feel you happen to have guessed the 'right' answer the DM had prepped for, and feel like his pawn (ie you've confirmed his creativity not your own). And if the DM says 'you don't know', then you're left feeling the DM has dismissed your clever idea and you need to look for a different 'right' answer, so your own creativity is again rejected.

    Psychologically, I think having your own creativity rewarded by objective, random dice is better than having it rewarded by a person. Likewise having it rejected by neutral - and random - arbiter is more comfortable too.

    Of course others might feel differently!

  12. In a lot of the examples posted here, the Knowledge skill really isn't that different from the Open Locks skill. It represents two things: A way to distinguish one character from another (i.e., "I fight with brains, not brawn!"), and a way to bypass challenges.

    When the party comes to a locked door, a successful Open Locks roll makes progress easier. A failed Open Locks roll means you have to look for either another way to get through the door, or a way around the door. Similarly, when presented with a need to know a fact, a successful Knowledge roll means you can keep on trucking. A failed Knowledge roll means you are stuck with a mystery on your hands, and need to find someone to solve it.

  13. Has anyone here played FATE? In those, a player can state a fact about something they have knowledge in and then roll, with the difficulty based on how likely the fact is to be true. If the dice roll succeeds, then the players statement is true, otherwise it's not.

    As an example, a player could say "I know a ritual that can bring my horse back to life!" He then rolls for a knowledge skill, maybe Arcane Knowledge. If he succeeds, he knows how to complete the ritual. If he fails, he's mistaken.

    This gives tremendous power to the players, and allows them to take the story in directions the GM wasn't expecting.

  14. Isn't this version just a skill tax, in which case it should devolve to a basic ability or class ability?

  15. @comments

    Some of this is starting to sound suspiciously like my original post. :-)