2e was D&D's maiden voyage into the wide world of skill systems that could actually be called such. I'm going to skip over a lot of its bizarre parts, like the way weird little subsystems sneak in through skills, some skills are blatantly better than others thanks to their combat applications... like I said, I'm skipping it. The point is that there's a separate skill for every field of endeavor that the designers could justify in a medieval-to-Renaissance-era magical world, sliced as thin as possible (Animal Handling, Animal Lore, and Animal Training are all separate skills). Later books introduce still more of these, which has the unstated (and, at the time, unrealized) effect of narrowing the applications of prior skills.
In a pattern that will be repeated for every tabletop game with a randomizing component, the character's knowledge of any particular fact or ability to answer any particular question is based on a die roll like any other skill check. In 2e, difficulty isn't really modulated off of the task in question, but off of the skill itself, as in most roll-under systems. The difficulty can
be modified with bonuses or penalties, but the game doesn't really emphasize that. Also, much like in D&D Next, a character's native ability (from ability scores) hugely outweighs training as a factor in overall competence in skills.
While I'm in the midst of a lengthy digression on 2e's skills, it also presents an alternate rules option that is a randomized approach to assigning a background, and decides you don't really need any rules to attach to that. The weird thing about this is that it's in the Player's Handbook.
The specific case of Religion is an outlier among skills. Sure, they all
introduce strange little subsystems, and that's... great... but Religion tosses out everything else
about how knowledge skills work. Your first proficiency slot gets you basic access, while slots beyond the first either improve your skill value by 1, or
you can gain the ability to check for knowledge about the religions of a broader region, or
you can gain such familiarity with one religion that you no longer need to roll a check to answer questions about it.
Third Edition D&D
I realize most of my readers are thoroughly familiar with 3.x and don't need this summary. It's included here only for completeness. Knowledge skills are the equivalent of other skills, and the rules are just as stingy about handing out Knowledge skills as class skills as they are with anything else. The game defines ten Knowledges, noting that the list is expansible with DM permission. Each of them grants a synergy bonus to another skill as well (or turning undead, in the case of religion). Intelligence is a double-dip benefit to Knowledge skills, as it grants skill points and an ability score bonus to all knowledge skills. This has the regrettable outcome of making clerics, paladins, and so on terrible at Knowledge: Religion, since they need to spend skill points on other things also. It is a grueling, uphill march against the system to play a fighter or rogue with significant ability in knowledge skills, all told. Even for a high-Int character, though, a list of ten knowledge skills to purchase individually means that probably
no one character has every Knowledge to any particular degree.
Early on, the value of a skill might
be mostly from your ability score bonus, but at higher levels, most of the skills you care about get most of their bonus from training. Also, the system supports racking up huge skill scores, and thus scales DCs to challenge the experts at high levels (a fact that directly informs 4e's skill design). It also has the result of pushing non-specialists out of the action.
Fourth Edition D&D
4e narrows the skill list drastically, but Arcana, Dungeoneering, History, Nature, Religion, and Streetwise remain as knowledge-gathering skills. I do like the (little-used) feature of knowledge skills that identifies monsters and their abilities - maybe if the rules gave you more reasons to specifically care about a monster's keywords. What I find interesting about this is that very few systems lay out such specific guidelines of the information a check grants. This is also one of the only cases where that kind of breakdown would be a good idea.
Intelligence is still the ability score of record here, though it doesn't double-dip on usefulness. Most
fighters in 4e won't have a knowledge skill other that Streetwise, but it's a lot more practical to do so since it only
costs a feat. Of course, 4e's approach to ability scores means that fighters dump Int as hard and fast as possible - but I talked about that a long time ago, in this post
. Knowledge skills don't include any internal specialization, and in general 4e gives the impression of preferring to gloss over mystery and lore. A serious analysis of why this is true probably deserves a post of its own, but this isn't going to be that post.
New World of Darkness
Moving on from D&D, nWoD also has a limited list of twenty-four skills. Because the system supports freely associating any possible skill/attribute pairing, you could theoretically make anything a lore skill by using Intelligence, comparable to the Rich Skills model
. I expect most Storytellers instead look to the eight Mental skills for lore functionality. The system also supports specialization within each skill, though one additional die for specialization isn't all that amazing. Ultimately, this part of the system is high-functioning, but it suffers a bit from other aspects of the game - players don't have enough incentive to spend precious skill points on their mundane skills, when awesome supernatural skills draw on the same pool of points and get the job done a lot
more decisively. (There are aspects of the system, such as rotes in Mage
, intended to address that; they don't quite go far enough, but it was a good idea.) In my own experience, I could have done a lot more to draw on other Mental skills, but the players absolutely got their money's worth out of Investigation and Occult skill ranks.
The other problem is one that plagued my players: successes only on 8, 9, and 10 mean that even very large dice pools often come up empty. This is the recurring problem with systems in which players roll the dice to see if the GM answers their lore question: should a character's knowledge be random? Does the excitement of a good roll on an obscure question outweigh the disappointment and undermining of a very poor roll to answer a very basic question? (Best-practices GMing includes not asking for rolls when a failure is pretty much nonsensical.) There's something to be said for taking the good with that bad from the dice, but I'd like tabletop designers to start considering knowledge as something less random. Anyway, I'm far from done with tabletop systems:
Legend of the Five Rings, 4th edition
L5R 4e (I have zero experience with earlier editions) has a large skill list, though far from the largest I've ever seen - that honor goes to either Alternity
or Skyrealms of Jorune
. Lore is one of these; it is in turn a Macro-skill, which I think
means that you advance in each sub-type separately rather than declaring and exercising an Emphasis. Overall, this is more punitive to the highly Lore-focused character than D&D 3.x, but far less punitive to the fighter who wants to dabble a bit. The system used here is otherwise unexceptional; the entry for Lore treats it as a curiosity that someone might bother with (but probably not). Ironically, the What Skills are Useful?
sidebar lists five things that are almost always useful for new players, and two of these are Lore skills. This may mean that most GMs really do make Lore skills central to L5R, and the text just presents it a little oddly. In looking at the various groupings of School Skills, it looks like a lot
of schools get a Lore (including, of course, nearly all of the Crab Clan getting Lore: Shadowlands... unfortunately a Low skill). This is another sign that I'm mistaken, and that's all to the good.
Trail of Cthulhu
Trail of Cthulhu
is famous for its unusual approach to skills, in that skills used to gain clues are not rolled as checks - if you have points in that skill, you spend them and receive the clue. This is a good way to avoid an adventure stopping cold from poor dice rolls. It does mean there's no tension in that phase of the adventure, and without tension it's open to question whether the investigation feels like the kind of challenge that the source narrative wants it to be. I wouldn't say this for most designers, but for something Robin D. Laws wrote, I'll take it on faith that the rest of the book goes into some depth on how to resolve this. My guess is that there's some tension in whether the team of investigators has enough of the right kinds of skills, and with thirty-nine different investigation skills, even a sizable and wisely-constructed party may have to think a bit on how to apply the skills they have to the problem at hand.
With seventeen different academic skills (plus some other skills that straddle the Academic/Technical line, like Art and Pharmacy), they hardly need further specialization, and it's hard to imagine some of the skills seeing use in even one adventure out of three. Accounting, Law, and Physics come to mind - though I tip my hat to any GM that proves me wrong. If anything, the game has so many knowledge skills that it approaches the design problems of player-defined lore skills - figuring out how to make a huge range of things feel useful. The text does discuss the inherent inequality of Investigation skills, and as long as the players read
the text, they won't get caught off-guard.
Yeah, I could go through a lot more systems, but the games on my shelf don't have a lot of things to add to the conversation. What I actually want to see in knowledge skills is support for both strongly lore-focused characters and characters who only pick up one or two areas of lore. I want games to know when to treat lore skills as bonus exposition
and when to treat them as extra clues
to solve a specific puzzle - that is, to make a hard puzzle easier than it was perhaps intended to be (in a way that feels rewarding rather than simply anti-climactic). I write a lot of DtD's puzzles, so I know how hard it is to write a puzzle with a default number of clues that is solvable, but with one or more additional clues is still an enjoyable challenge. (Making it solvable only with
the use of character abilities is, from my point of view, cheating on Plot's part, and running the risk of a ruined puzzle challenge when the wrong skill set shows up.)
Ability scores have a place when it comes to knowledge - it would be ridiculous for Intelligence not
to factor into knowledge in some way. In a class-based system, there's a definite expectation that some classes will need some ability scores particularly, allowing them to invest fewer points in the other scores. Fighter-types tend to get shut out of a lot of non-combat scenes because comparatively
few non-combat challenges focus on Strength (making an compelling physical-challenge sequence for a party of adventurers is also hard), and even when the class has something else to work with, they're so invested in Strength that their skill scores in other tasks are subpar. (Other classes than just fighters have this problem, but nowhere is it so consistent or absolute.)
I'd like to suggest that knowledge skills represent a mostly passive expression of Intelligence. They're some of the only core skills that players might sit at home and expect to use, continually asking the GM more questions until they come to a dead end of failed rolls or the GM's patience. Research, Deductive Reasoning, and Memory might be good replacements for knowledge skills in skill challenges. Knowledge skills, in turn, go back to something like the "fields of lore," probably using the three-level system from DtD. The three levels are usually characterized as high-school level, bachelor's degree, and post-graduate levels of knowledge, or something along those lines - even at the highest levels of lore, there are
still things you don't know. If you're just working on answering questions, you don't roll dice - the GM just decides whether or not the question falls within the purview of the player's level of lore. If you're in a skill challenge, the skills you're rolling are the verbs - Research, Deduce, Recall, and so on - with a "tool" bonus from your fields of lore.
Working out how to sell these lores to the players is the other big question, but that has to vary by system. For D&D Next, I'll suggest that each background should give a player one or more lores from a list - maybe that list is only defined by a few themes, and lore skills are otherwise player-defined. Some classes grant a fixed list of lores (such as wizards getting arcana without having to also choose an arcane-directed background), while other classes grant "one extra lore from your background." At first level characters have only basic lores, but these advance through... some means, possibly downtime action, over the course of play. Or maybe lores alternate with skills in their advancement?
Anyway. This probably needs more work, but it's a starting point.