I liked the idea of ability checks, where a player rolls a d20 (or 3d6) and if the total rolled is less than or equal to their ability score, they are successful. One of the things I like the most is that it makes every point of an ability score significant, and someone with a 9 INT is different than one with a 12 INT.
However, there are two parts of that classic rule that bother me.
The range of numbers is too wide, and each ability point is too significant
Not a fan of roll under mechanics, especially when nearly all other rolls are roll high. “So all night I have been hoping for 20, but now I want low?”
So I came up with the following ability score check breakdown, where each ability score now has an associated Throw value, which is the number that must be equal or exceeded on a d20 to succeed in some task. In addition to flipping it back to high rolls are good, it also flattens the curve while maintaining that each possible ability score is unique and different in some way*
The GM should feel free to apply a +/- of between 1 and 4 to checks that are more or less difficult.
Score Modifier Throw
3 -3 16+
4 -2 15+
5 -2 14+
6 -1 14+
7 -1 13+
8 -1 12+
9 0 12+
10 0 11+
11 0 11+
12 0 10+
13 +1 10+
14 +1 9+
15 +1 8+
16 +2 8+
17 +2 7+
18 +3 6+
- One might contend that I broke the rule where each score is unique
because both 10 and 11 have the same bonus and throw. However, it
has been my experience that scores of 11 are most likely to be
canibalized to boost your prime requisite by +1. In this fact 11
gains its uniqueness. If it bothers you, you can certainly change
the progression, but I liked the nice balance of 50% success in the
middle and 25% & 75% at the two extremes myself.