Telecanter had the idea of using die pools to the determine success of exploratory tasks:
So, if you are looking for traps/secret doors 6 means success (I roll high), normal characters roll 1d6, Elves roll 2d6, Dwarves roll 2d6 if it involves stonework.I like it. I like it a lot. I think it really slows down the advancement progression, which I felt was a little too speedy in my first write up and like how this method uses two variables instead of one (rather than having just X chance in 6, multiple dice are also rolled).
Take detect secret doors, for instance. Assume most characters roll 1 die and must roll 1 on that die to successfully discover a secret door. Thieves add one additional die to their pool at 1st, 5th, and 9th levels, increasing their chance of success. Elves, on the other hand, may roll either a 1 or 2 to discover a secret door, also increasing their chance of success. Therefore a 5th level elven thief rolls three dice and succeeds if one of those die lands with a 1 or 2 face up.
The amount of dice rolled represents increased skill, while the result(s) necessary for success represented the character’s natural skill. In that light, a character’s ability scores could be utilized. A character with a high intelligence, for example would be able to discover a secret door on a roll of a 1 or 2 (or 3 if the character is elf). Same assumption can be applied to strength for forcing in doors, dexterity for surprise, etc.
There are a couple issues, naming the amount of dice required and the additional trouble of having to reference both the party's marching order and the character's ability and chance to detect secret doors. A little more complex, yes, but a lot more detailed and mutable.