I think the biggest difference I see is that I try to reward PC skills instead of player skills. I see overcoming obstacles that require skills as being the responsibility of the PC, not the player. Several classes and builds have their contributions to the game tilted towards non-combat skills. I want to let people play to the strengths of their PC.
Nope. The player does provide the goal and approach, sure. But, if that approach is uncertain, it's the character that dominates. So, if the player isn't describing approaches that match his character's abilities, a bad time will be had when I call for checks, which turns out to be pretty darned often.
This is never 100% possible of course. Someone who is better at tactics will in general see a better resolution of combat. Someone who is better at solving mysteries and picking up on clues is going to be better at that aspect. I try to balance it out with appropriate skills, whether that's investigation, medicine, survival or whatever I or my players think up.
Funny thing, the tactics thing is right on because D&D strongly makes combat a challenge for players, not characters. But that mystery thing? I don't really have that problem, because the mysteries in my game aren't written out -- they're problems that the players propose solutions to. So, the character abilities feature very strongly in mystery resolution, and player ability to guess mysteries rarely helps because, when they fail a check, the mystery quite often heads off in a new direction.
That and I have no issue with how players state what they are doing. "I make an investigation check to see if the door is trapped and get a ___" is just as good as "I closely inspect the door looking for traps". It's just shorthand that speeds up the game.
I guarantee my game runs super smooth and fast when dealing with untrapped doors. Example from recent play:
Me: The hallway ends in a door that is a plain stone door with a simple handle in the center, much like all the other doors in this complex.
Player: I'll open the door, ready for danger on the other side!
This happens because my players trust my narration, and two, because I telegraph things that PCs are looking for, like dangerous traps. For a trapped door, the PCs get this:
Me: The hallway ends in a plain stone door with a simple handle in the center. Your light glints off a strange, shimmery green discoloration on the handle. What do you do?
Player: Hmm. I'm going to inspect the handle without touching it - does it smell funny?
Me: Okay, make a DC 15 perception check to try to identify the substance on the door.
Player: I'm trained in poisoner's tools, does that help?
ME: Absolutely! You recognize the substance as Thieves Bane, a contact poison. It can be neutralized with alcohol.
Player: I have some wine, I'll pour it on the handle.
Me: You wash off the substance, rendering it inert, what next?
Player: I'll open the door, ready for danger on the other side!
I don't hide the game.
So that's it. I generally run a very heavy RP game which has little to do with how I balance player vs PC skill. It in no way means that player decisions have no effect on the game, it just lowers the impact of player skill on specific aspects of the game.
I can't parse this very well, because it seems to encode a few assumptions of what "heavy RP" and "player vs PC skill" are. Since my players are currently in a large, trapped, temple complex, I'll go back to the session before they got in -- we had four extended social encounters where one group of PCs tried to improve relations with a faction (failed), solicited help from another faction (success), went to a prison to intimidate a witness into giving up a location (barely succeeded), and the other tried to track down a bad guy by talking to a third faction who pointed them at a contact (success), and then tried to intimidate that contact (failed). They all got back together and proceeded to track down a murderer and his gang, which had a social encounter that failed at the start (which I didn't expect but the players started, leveraging the limited fame of the dwarf in the pit fighting circuit to lull guards and try to get in to see the leader, but the dwarf absolutely bombed his CHA check) before having a big, multiple group fight as they went through the whole gang. Lots of RPing, lots of skill checks, where success and failure changed how the situation went. Want to know what I had prepped? None of it. Players surprised me at most every turn, but using goal and approach and assigning a consequence to failure, I was able to easily navigate the scenes. If the players succeeded, they got their goals and we moved to the next part. If they failed, they didn't get their goals and had to do something else. No magic words in sight.