So I'm hacking out a rough copy of a mechanics core that I call the Relativity Engine. There are two things I'd like to discuss on this thread:

*Does the concept of relative stats make sense (details below)?
*Is "community-building" a valid design goal, or is that something that lives on a different level than system?

Relative Stats: So the concept here is that a Relativity Engine game is concered with a group of characters who are all good enough at something to be doing it regularly or for a living, and differentiating between them based on their relative capacity at narrower area. For example, a group of mercenaries are obviously good enough at camping, fighting and getting contracts to survive as mercenaries. I'm concerned with whos better at full-on assault, whos better at strategy, whos better at horsemanship, etc.

This has two mechanical effects:

Any challenge that the party faces is scaled to the party itself - that is, mooks are going to be worse than most of the players at fighting, but the big bad boss at the end will be as good as or better than them.

Any one characters improvement knocks other characters stats down. Guy 1 was average with a sword, while guy 2 was better. Guy 1 improves, and gets better than guy 2 - now, in the group, guy 2 is average and guy 1 is better.

So does this make sense? I mean, if you got these concepts out of the rules, would it make you go "Neat" or "Huh?" (or whatever).

Community-Building As A Design Goal: So, I have a core resolution and character-creation system. Other system in a specific game built on the Engine would be created by the use of Mods - basically packages of rules about specific things (like combat, magic, car racing, drowning, etc). Now, ideally, I want other people to come up with their own mods and games, and every so often I would collate them and publish them (with full credit and everything).

This would necessarily require a community of people to read the core and be inspired enough to "finish" on their own. I want to support this, but is this really a design goal? Are there ways to do this in the actual design outside of just including the words in the text?

As always, any other thoughts that come up are welcome.

If you're looking at stats relevtive to each other, might I suggest a three teir instead of two?
Its on the tip of my tounge but I'm not too sure how to explain it. Just that the example sticks in my mind "Guy 2 might still be better than Average even though Guy 1 is Better than him..."
Might make more sense to myself I guess if there were a few other established group members...~

[Edit] Hit me just as I hit the "submit"- Average, Same, Better. Guy 1 can be Average, with some work become As Good As Guy 2, and with more work can be Better.