I was thinking about Diogo Noguiera's now-famous post "How to Never Describe a Dungeon!"
If you haven't read it, you might like to. He makes a persuasive argument. I'll summarize.
There are two ways to describe a dungeon hallway. One way is bad and, you should never do it. The other way is good, and you should try to do it every time.
The bad way is like this: "You get to a intersection and there is a door to the north and two passages, one going east and one going west."
The good way is like this: "If the corridor in the east leads to a natural cavern covered with mushrooms and myconids, maybe when the PCs look down that passage they will see a dim fluorescent light that emanates from the weird moss that lives there, and feel a light cool breeze flowing from that direction. Some moss may be growing in that corridor also. If to the west there is a nest of giant spiders, that corridor will certainly have more cobwebs covering it than the other passages they have been through, and some of them are still vibrating, as if something alive is touching the web."
The bad way is bad because it prevents the players from making an informed decision. With no information to base their decision on, they might as well roll the dice to decide which way to go. They might as well be in a straight railroad.
The good way is good because it supplies information that lets players make meaningful, informed decisions. Rather than a simple toss-up between "left or right" or "heads or tails" they have a real choice between moss and cobwebs, and whatever each implies about what's at the end of each hallway. (I would add, this approach also encourages players to use the information available to them to try to draw conclusions, and it rewards them for thinking ahead, by hopefully supplying them with a safer or more favorable path through the dungeon.)
I feel like Diogo has persuaded me, but what I'm thinking about is how to make it easier to put his idea into practice. (I say "his idea" even though I know Diogo's not the only one who's ever made this point - he IS the one whose essay on this point I'm quoting here.)
Because good intentions are fine, but I think there's a reason so many game-masters probably default to the bad way of doing things. The reason is, it's easier.
Pick any dungeon you like. Open the map, find a hallway. Imagine your players entering the hallway. Now, prepare to describe it to them. What will you do?
The first thing you'll do, of course, is look at the drawing of the hall on the map. How long is it? How many doors open off of it? Where are those doors? Are there any special features you need to mention?
All of that is important information, but if that's all you describe (and I'll be honest, it's probably all I usually ever describe) then you're doing things the bad way that we just agreed doesn't allow for meaningful player decision-making.
What more do you need to do, in order to describe this hall the good way? Well, you need to take note of the room numbers. You need to go check the key to the map for each of those rooms. For each room, you need to read the description, come up with a general impression of what's the most important thing in the room, decide what a good clue about that thing would be, and then move on to the next room, which is probably on another page. Also, you need to hurry, because your players are waiting, and what's taking so long describing a simple hallway?
Did I say the bad way was easier? I misspoke. The bad way is MUCH easier.
So if all you have is good intentions and a regular map, it's going to be much harder to describe each hallway the good way. And that's because RPG maps don't number hallways.
Which is one of those things that I noticed, and thought was odd, when I first started reading RPG books. But there are all kinds of conventions and practices that go into making RPG books. Pretty quickly you get used to them, and basically forget that you ever thought they were strange. But reading Diogo's essay made me remember the way I felt when I first looked at an RPG map and wondered why the hallways didn't have numbers like the other rooms.
Which is fine, I guess, if you're treating hallways as non-spaces, like we've probably all been doing all along. But treating them that way basically forces you into the bad way of describing halls, because it forces you, the GM, to come up with a description right there on the spot.
If we want to describe hallways well, we need to number them on our maps.
Every hallway is an empty room, except we usually don't think of them that way. They have all the promise, and all the problems, that we associate with empty rooms. Think of every blog post, every essay, you ever read about what to do with empty rooms - we should apply all those ideas to hallways. And we should number them.
Getting good hall descriptions during play shouldn't be a matter of exception skill, fast reading, and quick thinking. It should be a matter of looking at the hallway's number on the map key, and communicating the information there to your players.