Building interesting dungeon rooms

A dungeon crawl which consists of a series of door-monster-treasure encounters can become pretty boring. More interesting dungeon rooms present the players with some decision to make. Now we all probably encountered rooms of the type "You enter a room with a red button on the wall. (If somebody presses the button a fireball explodes)." That is a decision, but not a very fun one. You can't find out what is the good decision without making the bad decision. So I am looking at rooms which answer the following design questions:

• What do the players see? (Example: A room with an exit that can't be opened by normal means. Left and right of it are altars with statues of two different gods, with censers before them. In a nearby chest there is some incense.)
• How can the players find out more? (Example: A religion check, or looking through books in a library the players passed earlier, reveals that one of the gods is the god of light, the other the god of darkness.)
• What is the bad decision and it's consequence? (Example: Lighting the incense in the censer of the god of darkness opens the door, but also casts a spell of darkness on the room. The players are then attacked by shadows from the room behind the door, which have advantage due to having blindsight.)
• What is the good decision and it's consequences? (Example: Lighting the incense in the censer of the god of light also opens the door and releases the shadows on the players. But it also casts a spell of permanent light, which gives some disadvantage to the shadows.)

The idea is that the room doesn't have to be overly complicated, nor designed to lead players into making a mistake (no "Grimtooth's Traps"). It just should make the players stop and think for a minute, and reward them if they solve the mystery.

I would like to challenge your creativity. What interesting dungeon rooms can you think of, which you can describe using the above template?

I head in an opposite way. For my table, long dungeon crawls for the sake of a dungeon crawl isn't an interest. I usually plan out my dungeons using the Five Room Dungeon method (google for more then you ever wanted to know). It doesn't refer to five actual rooms, just five main points of interest:

Room 1: Entrance And Guardian
Room 2: Puzzle Or Roleplaying Challenge
Room 3: Red Herring
Room 4: Climax, Big Battle Or Conflict
Room 5: Plot Twist

Now, onto that I'll add in additional points of interest for plots and campaign arcs, including laying foundation for ones that characters aren't aware of and points that would make sense for other campaign arcs then than what the PCs think they are doing at the exact moment. "Onto" could refer to more locations, or overlaying them on these locations. If the primary foes are summoned demons all wearing a fine iron chain inscribed with tiny ruins, that's an easy link into one plot and should make them wonder why that diabolical patron is involved, etc.

I like color for color's sake, but I also don't want to draw out a dungeon for too long. I'd much rather gloss over "searching and finding several empty rooms - bedchambers and the like, covered with undisturbed dust and long-ago ransacked" then to map out going through each of those rooms and making each on have enough details that we spend 10-15 minutes each and eat up an hour of the 3 hours every other week we play without having gotten advanced the plot. An hour of play is about the equivalent of five days of waiting for the next session - I want to make it count.

That said, there are plenty of quickly resolved tools I can use to make a big complex feel, well, big. My #1 tool, taken from 13th Age, is the montage. Once the players have the feel, I'll do a montage scene of exploration, with each player contributing a minor challenge or puzzle and how they overcame the last one.

Another I use are challenges, which in some ways are free-form descendants of skill challenges but without the formal "X successes before Y failures", also allowing use of resources (spell slots, consumables, knowledge, etc.). I might also use a montage to show how successful (or not) they are.

Let me give an example of that while combining with not taking the rooms literally. In one "dungeon" the Entrance and Guardian "room" for a Ice Castle made by a Vampiric Frost Giant Ice Sorceress was how to find the throne room they knew they were heading for while avoiding ice-based traps and Frost Giant Zombies. But in a more abstract way, not as a series of encounters, sicne ay particular trap or small group (heh) of Frost Giant Zombies was too low a level to kill them.

So we set up about the rogue taking point, how to efficiently deal with the Frost Giant Zombies (once they discovered they weren't an on-level combat challenge), spells, methods and skills used to make as direct a bee-line to the throne room in the purposefully confusing castle to minimize the traps and frost giant zombie encounters - then it went to a montage of dealing with them based on it, including damage and resources used, to get to the throne room. Took 20-30 minutes total.

I'll just set a nice baseline, so succeeding respondents can get more fantastic from here:

What do the characters see?
A dungeon room. A torture device is a good start...the Rack is set up near the wall, with a thick pile of straw laid out beneath it. Its operator sits nearby, and PCs can't help but notice a shiny bell resting at his foot.

How can the players find out more?
Talk to the torturer. Deploy animal companions.

What is the bad decision and its consequence?
Anger the torturer. Attacking him is a good way to do this. His response is to ring his bell and run, alerting the next 10 chambers that there are enemies nearby.

What is the good decision and its consequences?
Bribe the torturer, or offer up the party bard as a sacrifice. Then he'll let PCs go peacefully. And the bard...piecefully.

I don't know if [MENTION=6688858]Libramarian[/MENTION] still posts on these boards, but he used to have good ideas for this sort of thing.