Keep Those Dice a'Rollin'

So my friend Jonathan came out this past week (Huzzah!) and we got to game a little bit. I am still fairly enamored with my new experience system, especially how clean it looks/feels to me at the moment. One potential concern that I noticed is that my system for saving throws and skill checks may be inherently flawed. The concept is that you roll 3d10, and you are aiming for a number lower than the relevant stat with a bonus gained from a secondary stat. With an expected average stat of 15, and a range of 3 to 30, I figured 3d10 would be appropriate. However, it seems that the majority of rolls are landing above 16 even after adjustments. I should have expected this, but I'm not yet sure how to deal with it. I was trying to approximate a d30, but I don't actually have any of those and they're not terribly common anyway. I suppose I could do a d6 and a d10, like how d20 rolls used to be done before d20s became popular. 1-2 on the d6 means the d10 is 1-10, 3-4 means it's 11-20, and 5-6 means 21-30. I'll have to do some blank rolling to find out if that brings the average down.

In pondering further upon the topic of skills, I thought about the idea of a slow method of improving skills based directly on Intelligence, rather than class as 3rd Edition had it. I'm thinking about putting a limiter, though, on how much of a bonus you can give any individual skill, based on the primary stat governing that skill. Maybe half of the stat? One of the things that really bugged me about 3rd Edition skills was that within a few levels the bonus to the skill far outstripped the ability of most challenges to keep up. Difficulty ratings kept getting arbitrarily bumped up just to have any chance of failure. I'm also, personally, a believer in inherent ability being a greater contributor than training.