Design Concept: Specific Skills, Not Broad Skills

I'm taking some heavy influence with this statement from the OSR (Old School Renaissance) community and the concept of "Your Sheet Has No Answers". The idea of specific skills revolves around the discussion about where is the line drawn between the character and the player. If you look at some modern game designs such as 5e, Pathfinder, or even Call of Cthulhu, the character sheets are filled with skills and abilities. In my experience players will often approach a situation, assess it, and then ask their GM if they can roll some dice to gather information or do a task. 

That approach to game design works pretty well. It ensures that the character in the scene is acting based upon the skills they have and that their player might not. For example, if I were playing a traditional cleric, I would presume to have in-depth religious knowledge about the game world. As a player I might have very little knowledge of the in-game religions. To expand further, I might not be a religious person at all to even have the proper frame of reference to a apply those concepts to a character. 

What I don't like about these systems is that they often have the player "give up" their investment into the scene too early. Perception checks in D&D are one of my most frustrating and hated mechanics. Too often my players walk into a room and immediately ask for a perception check. Checks for skills in general are good and should be used in games, but when you have certain checks dominate scenarios you then have a poorly designed system. Perception checks themselves are too broad and too encompassing. How many times in a game are you making a perception check versus a nature check for example?

Instead I would like to see players describe to me what they want to do in the room, what they are looking for, and how they want to look for it. It doesn't have to be specific. They can state that they are looking for traps, or they want to smell the air for gun powder, etc. But, if they ask the right questions, they shouldn't even need to roll anything at all. I can give them the answer.

The perception check problem rears its head again when players ask to roll and fail. Let's say the party enters a room and they need to find the hidden door down to the dungeon. If they roll a perception or investigation check and they all fail, the GM is now left having to manage how to keep the story moving forward. Do they convey any information? Do you have the party roll again? If there is no failure state with risk the roll becomes completely useless. 

How do we fix this then? I feel the answer is specificity. Create roll situations that gives information that only that character can obtain. Create abilities, checks, skills, traits, etc. that are specific and unique to the character and can be applied to specific situations. The payoff then is much more rewarding. If for example, you walk into a room and you have a chef in your party, they might have a skill to specifically smell the air or identify ingredients in a dish, or know why that wine's color is slightly off because of the poison in it. 

Specificity creates a much more rewarding experience at the table, where everyone playing gets a moment of enlightenment and excitement because the right tool was used at the right time.