Legend & Lore: Player vs. Character

When it comes to skills I greatly prefer the approach of later editions: let the player tell you what they are generally trying to do, and then let a combination of dice and the character's modifiers to determine the outcome. Mearls argues that this approach detracts from immersion because it causes players to focus on their sheets rather than the environment. I would argue that it actually helps maintain it because characters good at stuff that the players are not are still likely to succeed, and things that the player is good at but the character is not are more likely to fail.

In other words it helps sustain the players' beliefs that the players are pretending to be someone that are not themselves.

On the topic of the environment, in past editions I guess the players had to engage in a combination of 20 questions and Zork in order to find...whatever it is they are trying to find. Personally I started with Basic and recall having to make ability score checks to do stuff like snatch food out of Axel's hand or keep him from getting pissed off but, eh. At any rate the attention to detail is utterly irrelevant to the mechanics used to interact with it. If I wanted to I could describe all the furnishings in a room, and even allow players to state in excruciating detail what they are doing with it. My main problem with this route is time.

Maybe in some games players enter a room and declare that they are going to search it and just make Perception checks--which is fucking fine by me because it strips out the endless torrent of queries--but usually in my games players actually call out specific things in a room to investigate, in which case I go into additional detail and/or call for a skill check if appropriate. Here I find that being able to make a Perception (or Dungeoneering, History, Arcana, etc) check to indicate important details and/or secret shit a huge time saver, as rather than going down a list of search criteria I can encapsulate it all with a single dice roll and more accurately simulate the character's ability to search and find things--often resulting in the party missing out on important things much less often.

Finally, I feel that this approach helps level the playing field between newer and veteran players. As Mearls states in the article, older editions required the player to actually learn information about monsters; the cited example was the weakness of trolls, but I recall that OD&D made it so that you had to try charm spells on monsters until you figured out which ones it worked on. The problem was that once the player figured this out, she could either use it in future games with other characters, or have to pretend not to know and "stumble" upon the information (probably in a contrived manner). In this instance immersion is reinforced by the fact that the book-worm character has a chance to know this sort of thing instead of figuring it out via trial and error.