If you've been a regular reader of the blog over the years, or even if you've just read a bunch of the posts about game design that I've made in that time, you'll note that I strive for elegance in systems. That doesn't mean striving to make simplistic rule sets, but it also certainly doesn't mean introducing complexity for the sake of it. I like sturdy, multifunctional rules that apply a common method of resolution to a variety of situations while feeling natural in those situations.
I like the idea of a simple (6 and under)/(7-9)/(10 and over) mechanism like the one found in Apocalypse World, and for some tasks it feels natural, but in my experience it finds it's limits quickly. This is especially true when it comes to the variant difficulties of tasks, or the variant skill levels of characters. The way that people tack new sub-rules onto the base to accomodate these issues indicates to me that the core rules were never meant to work in certain ways, and shoehorning the core rules into those areas with sub-rule variations is simply inelegant design. I guess that's one of the many things that bugs me with Apocalypse World. It was meant to be a collapsible design structure, where you could forget bits of the rules, use core rules and it would still work fine... but hell if you're going to do that, why not just play something like Risus or Ghost/Echo. They even have a better way of handling variable difficulty factors.
While working on The Fen, I've strived for elegance while trying to avoid paths I've taken previously. I think the SNAFU system used in The Law is good for it's i tended purpose over there... but it doesn't necessarily feel like the best bet here. This game feels like it needs to be more procedural, self governing, almost with a crude AI to help determine threats and potential benefits, feeding in a variety of factors to give outcomes that are relevant to the tasks being undertaken, while avoiding needlessly "Rolemaster"-levels of complexity.
At first I was tinkering with a flat d6 system, where you might add an extra "positive" d6 or two if you had some kind of advantage, or an extra "negative" d6 or two if things were against you. Positives and negatives cancel out, if you end up with one or more "positives" left over you roll multiple dice and keep the best... if you end up with one or more "negatives" you roll multiple dice and keep the worst.
The basic result outcome would be...
1: Fumble. Something bad happens directly related to the action.
2-3: Fail. Nothing much happens.
4-5: Success. The expected outcome of attempting the task occurs.
6: Special. A critical success, where some kind of unexpected bonus is added to the outcome.
It feels simple enough, and the various tasks in the game, such as tending the fire, exploring, crafting, fighting creatures of the fen, etc., can each have their own overlay applied to the basic mechanism... so yeah, a bit like Apocalypse World 'moves', but since this is project is more akin to a cooperative boardgamethan an rpg, the regulated system generally feels alright.
Then I started adding more complexity to the system, where some character skills or situational modifiers added +1 to a die roll rather than an extra die. This eliminates the chance of a fumble, and doubles the chance of a special outcome (or makes the numbers go weird when combined with the multiple dice options). Then I decided that the 'special' outcomes should have six variations (roll a d6 at the end of the task resolution to see which bonus effect applies). Then I wanted potential result outcomes with more than 6 possibilities, so I thought about adding d10s into the mix. In other cases I couldn't even think of 6 possibilities, so I considered halving the result (1-2: option 1, 3-4: option two, 5-6: option three)... and the whole thing was starting to feel inelegant. It was starting to feel limethe kind if system where every action us preceded by looking up the rulebook to get a quick reminder of what you're supposed to do in each specific situation. The opposite of what I want.
I still like the core, roll d6, outcome based on (1)/(2-3)/(4-5)/(6), roll positive and negative dice if they haven't been cancelled out by other factors. It generally keep the game moving forward, which is a good thing as the creeping difficulty ramps up in the background. Now it's just a case of applying the right contextual inputs and outputs to that system without overly complicating things again.