I tend to get annoyed at players who almost completely forego interacting with the fictional world.
"I search for treasure. I roll a 20!"
<snip>
I make clear that a roll only counts after you tell me what you're doing
Here are some passages from Gygax's PHB (pp 24, 27; published 1978) and DMG (pp 19, 60, 97; published 1979):Gen Y gamers tend to want a dice roll for everything. For example, during our AL games, the 3e/PF players will say something like, "As I enter the room, I try to percept it. What's the DC I need to beat?" For an old fart like me, I'm thrown off. "What do you mean? Are you searching the room? How are you searching it?" And that in turn throws them off. They were just expecting a die roll and compare the result to a DC, and if they beat it, I as the DM need to tell them everything about the room.
<snip>
Another example is if the players know there's this monster X they will end up running into, the Gen Yers are "I rolled an 18 on my lore check, what do I know about it?" compared to the GenX players who will say, "I find a library or ask around to see how much research I can do to find out about monster X."
Rangers are a sub-class of fighter who are adept at woodcraft, tracking, scouting, and infiltration and spying. . . . Rangers . . . are themselves surprised only 16 2/3% of the time (d6, score 1). . . .
Secondary functions of a thief are: 1) listening at doors to detect sounds behind them . . . Listening at doors includes like activity at other portals such as windows. It is accomplished by moving silently to the door and pressing an ear against it to detect sound. . . .
Hearing Noise: This is pretty straightforward. The thief, just as any other character, must take off helmet or other obstructing headgear in order to press his or her ear to the door surface in order to hear beyond. . . .
In addition to the simple exercise of observation, many times characters will desire to listen, ear pressed to a portal, prior to opening and entering. This requires a special check, in secret, by you to determine if any sound is heard. . . . When a die roll indicates a noise has been heard, tell the player whose character was listening that he or she heard a clink, footstep, murmuring voices, slithering, laughter, or whatever is appropriate. . . . Failure to hear any noise can be due to the fact that nothing which will make noise is beyond the portal, or it might be due to a bad (for the listener) die roll. Always roll the die, even if you know nothing can be heard. Always appear disinterested regardless of the situation. . . .
[Secret doors] are portals which are made to appear to be a normal part of the surface they are in. They con possibly be sensed or detected by characters who are actively concentrating on such activity, or their possible location may be discovered by tapping . . . Discovery does not mean that access to the door mechanism has been discovered, however. Checking requires a very thorough examination of the possible secret door area. You may use either of two methods to allow discovery of the mechanism which operates the portal:
1. You may designate probability by a linear curve, typically with a d6. Thus, a secret door is discovered 1 in 6 by any non-elf, 2 in 6 by elven or half-elven characters, each character being allowed to roll each turn in checking a 10' X 10' area. This also allows you to have some secret doors more difficult to discover, the linear curve being a d8 or d10.
2. You may have the discovery of the existence of the secret door enable player characters to attempt to operate it by actual manipulation, i.e. the players concerned give instructions as to how they will have their characters attempt to make it function: "Turn the wall sconce.", "Slide it left.", "Press the small protrusion, and see if it pivots.", "Pull the chain."
It is quite acceptable to have a mixture of methods of discovering the operation of secret doors.
Secondary functions of a thief are: 1) listening at doors to detect sounds behind them . . . Listening at doors includes like activity at other portals such as windows. It is accomplished by moving silently to the door and pressing an ear against it to detect sound. . . .
Hearing Noise: This is pretty straightforward. The thief, just as any other character, must take off helmet or other obstructing headgear in order to press his or her ear to the door surface in order to hear beyond. . . .
In addition to the simple exercise of observation, many times characters will desire to listen, ear pressed to a portal, prior to opening and entering. This requires a special check, in secret, by you to determine if any sound is heard. . . . When a die roll indicates a noise has been heard, tell the player whose character was listening that he or she heard a clink, footstep, murmuring voices, slithering, laughter, or whatever is appropriate. . . . Failure to hear any noise can be due to the fact that nothing which will make noise is beyond the portal, or it might be due to a bad (for the listener) die roll. Always roll the die, even if you know nothing can be heard. Always appear disinterested regardless of the situation. . . .
[Secret doors] are portals which are made to appear to be a normal part of the surface they are in. They con possibly be sensed or detected by characters who are actively concentrating on such activity, or their possible location may be discovered by tapping . . . Discovery does not mean that access to the door mechanism has been discovered, however. Checking requires a very thorough examination of the possible secret door area. You may use either of two methods to allow discovery of the mechanism which operates the portal:
1. You may designate probability by a linear curve, typically with a d6. Thus, a secret door is discovered 1 in 6 by any non-elf, 2 in 6 by elven or half-elven characters, each character being allowed to roll each turn in checking a 10' X 10' area. This also allows you to have some secret doors more difficult to discover, the linear curve being a d8 or d10.
2. You may have the discovery of the existence of the secret door enable player characters to attempt to operate it by actual manipulation, i.e. the players concerned give instructions as to how they will have their characters attempt to make it function: "Turn the wall sconce.", "Slide it left.", "Press the small protrusion, and see if it pivots.", "Pull the chain."
It is quite acceptable to have a mixture of methods of discovering the operation of secret doors.
In light of these passages, which informed a lot of people's play of the game, I don't think there is anything very novel in the idea that a character might declare a perception check of some sort and than anticipate a roll (whether made by the GM, or by the player). The idea that this might be done for visual perception as well as for listening is already implicit in the AD&D rules for detecting secret doors, for a ranger's reduced chance of surprise, etc. There was never any expectation that a player had to declare how his/her ranger was scouting and/or spying so as to get the surprise bonus; or how his/her thief was listening so as to get the better-than-typical chance to hear a noise behind a door. And as Gygax says, either method for detecting the method of opening a secret door is acceptable.
Here are some more passages (PHB pp 27, 118; DMG p 20):
At 4th level (Burglar), thieves are able to read 20% of languages, and this ability increases by 5% with each additional level of experience until an 80% probability is attained. This enables the possible reading of instructions and treasure maps without having to resort to a magic item or spell. . . .
Read languages: This ability assumes that the language is, in fact, one which the thief has encountered sometime in the past. Ancient and strange languages (those you, as DM, have previously designated as such) are always totally unreadable. Even if able to read a language, the thief should be allowed only to get about that percentage of the meaning of what is written as his or her percentage ability to read the tongue in the first place. The rest they will hove to guess at. Languages which are
relatively close to those known by the thief will not incur such a penalty. . . .
Legend lore and Item Knowledge Percentage shows the chance that the bard has of knowing something about a legendary person, place or thing or of knowing what a particular magic item is.
Read languages: This ability assumes that the language is, in fact, one which the thief has encountered sometime in the past. Ancient and strange languages (those you, as DM, have previously designated as such) are always totally unreadable. Even if able to read a language, the thief should be allowed only to get about that percentage of the meaning of what is written as his or her percentage ability to read the tongue in the first place. The rest they will hove to guess at. Languages which are
relatively close to those known by the thief will not incur such a penalty. . . .
Legend lore and Item Knowledge Percentage shows the chance that the bard has of knowing something about a legendary person, place or thing or of knowing what a particular magic item is.
These passages indicate that the idea that a player might make a die roll to determine whether or not his/her PC has knowledge of something not as a result of actual game play, but as a result of the PC's "off stage" experience and learning, has an equally ancient provenance.
I would expect a 5e character built with a good INT bonus and trained in knowledge skills to play much the same as an AD&D thief or bard did, in respect of whatever the field of knowledge it is that the character has expertise in.