|I need to keep out of the sun for a bit.|
After reading through the Planet Eris house rules supplement, it occurred to me that 2d6 for skill checks for thieving abilities (or even other possible skills to be added later) has the huge advantage of the built in bel curve.
You can give everyone a basic chance to find traps or hide in shadows and the like, and yet give thieves better chances than other initially and allow them to increase with the bell curve.
So, say you give the basic chance to find traps as 11 or better on 2d6. That's about a 8% chance (doing the math in my head, so I could be very wrong). Give a bonus to finding for a high wisdom (+1) and / or being a dwarf (+1) would make that 9 or better - roughly 28% and never going up.
Now, if we start thief skills for thieves at 9 or better (Climb Walls and Read Languages would need to be tweaked) at first level, without adjustments for high ability scores or race, our thief is succeeding 1 in 4 times.
Give the thief 3 plus ones at 1st level and 2 plus ones at every level after that (no more than 1 plus to a particular skill at each level) and the thief can specialize, but as one gets better in a particular skill, the returns for that plus 1 get smaller and smaller. It's more effective to be a generalist, at least at higher levels.
A roll of 2 would always fail, no matter the skill level or the circumstances.
I need to really write this down and tweak it, but there is my thinking out loud, based on the Thief Skills charts presented in Planet Eris (I would move all of the 1d6 skills to the 2d6 resolution method).
Or this could all be the results of a mind suffering from sun poisoning...