The recent activity on the board has inspired me to write a little bit about our D&D adventures, again. I'm generally quite lazy about doing this, which is a shame, as the Internet really does a lot for the rpg hobby - even if I might not get that much personally from writing long actual play reports, other people do. (I know this to be the case because I've myself learned the craft primarily not from rulebooks, but rather from specific essays of the Internet age.) Besides, I often find it useful to have an old essay somewhere to point to when a topic comes up again.
I'm inspired to write about my current skill system, because I think it's pretty unique and clever. I think that this sort of mechanical stuff is very contextual, so I don't expect others to be able to use this as is, but there are certain insights here about percent-based skill systems and such that may be useful to somebody. This is going to be hardcore traditional task resolution logic throughout, but perhaps you'll find some saving grace in the fact that I stand on the shoulders of giants and may therefore have a better handle on this than some '80s writers wrestling with the same issues.
The following treatment has been in use for 49 sessions of play at this writing, so it's seen some actual play. It's the biggest change I've made to my homebrew rules since 2011 or so.
How we resolve tasks
All characters have the same matrix of Ability scores:
STAMINA - a generic measure of physical competence
WITS - how quick-thinking and alert the character is, generally speaking
KNOWLEDGE - how much lore and skill the character has
WILL - how in tune with himself the character is
CHARISMA - social intelligence, pretty much
Abilities are initially rolled 3d6 in order, but they have various ways to fluctuate during the campaign, they're not nearly as static as is normal for D&D. For instance, they have a chance to go up at the end of every session, and every time a character is injured there's a chance for them to go down.
When a task is in need of resolution, the GM determines a target number (DC, as one might as well call it - this is all very similar to 3rd edition D&D) and the Ability used to resolve the situation. The player rolls, adds their Ability score to the result, and compares to the DC. The DC for Ability checks is usually 20, but it is determined in realistic terms, so can go up and down depending on the situation.
OK, so that is stupid simple so far. Here's some nuances:
* A natural '20' explodes, so you roll and add again.
* A natural '1' is a "fumble", which is treated as an "extra degree of failure" in terms of creative interpretation of the dice. The result of the roll is calculated normally, which may, depending on the target number, mean that you get a success despite the fumble (which is basically a "yes, but" result), or you may fail as well as fumble ("double failure", sort of).
* Every 5 points of difference between the DC and the result of the roll is treated as an extra degree of success or failure, so +5 difference is a "2nd degree success", +10 is "3rd degree" and so on. These excess degrees of success or failure are used to essentially "purchase effectiveness" in a semi-formal way for various tasks.
The reader can see how this is relatively math-intensive when you come down to it; many rpg systems recoil from rapid addition and substraction of two-digit numbers. My only defense is that I'm pretty much designing for our local use, and math-aversion is apparently not as common in Finland as it is in some parts of the world. The success degree calculations ("I rolled 27 and the DC is 35, what's the degree of success?") are quick to become a second nature, I've found, and we don't really need to think about it in practice.
The success-counting is the most heterodox part in the system compared to traditional D&D, which generally rolls on pure pass/fail. It's also the most formalistic bit that ties the task resolution into greater concerns of success and failure, because the "degree of success" naturally defines a currency of sorts for measuring various procedural concerns: one degree equals one useful fact discovered; one more enemy caught in the blast radius; one less check needed later on; one more die of damage; one point of damage avoided; one day less of crafting; one more henchman mustered. The list goes on and on, but it's important to understand that although there are many procedural rules of thumb for converting and interpreting success degrees into concrete events in the fiction, there is no formal guarantee of their value; it is very important for an organic task resolution system like this that an easy success in a meaningless task may well produce a bunch of "degrees", but that does not mean that you can outright trade them for something useful in an unrelated matter. All this is merely semi-formalistic, not a closed formal system by any means.