For example, suppose we wanted a rule that decided a character's reputation. Because reputation seems to have little or nothing to do with any other number in the lexicon, it would seem logical to create a new number, something that applied only to reputation, in order to address this very obvious "lacking" in those things that define a player.
But is it a lacking? And is it really not represented in the character's stats? Does it not follow that a character with a high wisdom, charisma or intelligence is less likely to make enemies, annoy people, develop a bad reputation and so on? Does not the reverse also follow, as a character acts selfishly, stupidly or boorishly? Why not simply make the "reputation" an average of the player's existing stats? Why dismiss these points of reference in favour of creating a completely unrelated point of reference?
The answer is that most players are always looking for something that will mitigate the negatives on a character, giving them at least one more chance for a positive. My character's wisdom is 10, his intelligence is 9 and his charisma is only 8, but at least he has a reputation of 18!
Unfortunately, that reputation doesn't mean much. Because it isn't enfolded into the rest of the game (meaning there are no rules that define how everything relates to reputation like it does to hit points or movement), that number just gets ignored . . . until a year or so passes and the player asks the DM, "This reputation number that I have: do I still need to keep track of that?" Whereupon the DM admits, sadly, that no, that's no longer necessary.
I did break my rule, the one I started this post with, when I introduced the idea of Sage Abilities.
Initially, I did have those Sage Abilities exist as a completely separate entity. And it did not work. It was based on a percentage system of knowing something and upon "questions" that the players would ask to see if they knew something . . . but the questions weren't clear and the % chance of knowing was painfully inconsistent and for years the system just languished on the edge of death because there was no solution for it that I could see. Still, I needed something that would fix the problem of what do players know, so I patiently did not press the issue and waited for clarity.
The inconsistency of the die roll is the killer. This is the worst part of the "skills" system that was introduced with 3rd Edition. Seriously - why is it that my character can do this thing this minute and can't do this very same thing the very next time he tries? Why is everything based on odds? My own system tried to make it that way because I didn't want to try to codify all of human knowledge and then decide what players knew and what they didn't (though, to be truthful, that was sort of the inevitable solution).
At least there should have been a precedent system in place for 3e (maybe there was one, I never played that system). If my character was able to jump a 9 foot gorge once, then he should be able to jump 9 feet forever (unless some other extenuating circumstances existed). That could have been managed by keeping the character's roll for any given skill attempt, then counting that as a minimum roll and letting the player try to best it with each new opportunity. Any roll less than the minimum would mean no improvement. That would sort of make sense. But having a totally unique roll happen again and again for similar situations - that's way too inconsistent to build a thoughtful system around.
The change in the Sage Abilities when I began to work on them more seriously was the reasoning that these are not things the players should roll against, but things the players can absolutely do or things the players will absolutely know. The reason it is taking years to build the system is what I said just now: codifying all of human knowledge and ability. It is painstaking work, made more painstaking by having to constantly create rules for things that there are no rules for. How do I work out how much a character knows about horseriding when there are no rules for horseriding?
The true system, however, did not come from inventing a system with knowledge points that - like reputation points - had a chance of standing on their own and having nothing to do with the rest of the player's character sheet. The spark came when I realized that every skill and ability that every character class had - from monk open hand attacks to spell use to weapon proficiencies to a cleric building a temple - all FIT into the same immense umbrella of a system. Rather than make a stand alone system for Sage Abilities, I gathered in many other stand alone systems - turning undead, morale, training other people, creating magic items - into a single overall formula.
No one, I know, wants to do this kind of work. Hell, I may not live long enough to do this kind of work. But each time I add another part to it, another piece, it gets stronger and more viable in the players' minds.
They LOVE the knowledge numbers, not because its a chance for their characters to shine be better at something, but because they make something that was formerly hopeless and unclear into something precise and known. Truth is, if I sat down for a week or two and just made headings for the sage abilities, right through to bard and monk, the players and I could work out the rules for those things on the wiki as we moved forward from game to game.
Maybe that is what I need to do; instead of trying to trim proper rules, one by one, maybe I should start with all the categories in place and then let it be messy as slowly, steadily, we work each category out.