Experimentation and discovery: important and fun elements of the game. We should examine how we can deliberately include them in our games. Traps that only interact with your Saves, and Obstacles that only interact with your Skill Check, lack these elements, and are generally not fun. Let's do something else.
I want to define a broad category of things that I can't think of a good word for. I can, however, define their likely qualities. I suspect that you could break any and all of these rules and have the thing still serve the same purpose in the game, or have the thing obey every rule and yet not serve that purpose.
- They serve as an obstacle or impediment, preventing or dissuading access to something behind them (possibly metaphorically)
- Their operation is deterministic, or at least somewhat predictable
- They respond to certain specific inputs, possibly in the form of triggers, sensors, behaviours, or manipulable controls
- Experimentation combined with Thinking is usually the best way to find out how they work. Possible alternatives include getting the information out of an NPC or institute that does know, or expending resources (divination spells)
- Knowing how they work is usually the best way to cease being impeded by them, though one may be able to remove the impediment (temporarily or permanently) by expending resources; either personal things like spells, or equipment, or money, or time. Or, just by taking a risk.
- They provide enjoyment by being something to play with.
Things that are part of this broad category:
- Some puzzles (if you do them right)
- Traps (if you do them right)
- All manner of unique, possibly arcane, machinery in your dungeons, workshops, and wizard laboratories
- Most Black Doors
- Magic items (maybe)
- Golems (maybe)
- Bound devils/demons (maybe)
- A north/south passage that you can only walk through while facing south
- A 20' chasm with no bridge and the ghosts of 10,000 angry geese
- An immortal Skeleton Jelly that steadily splats wetly after you without getting tired
If all the relevant information of the obstacle is immediately obvious, then you get a closely related group in which the experimentation, discovery, and play exist only within the mental process of nutting out a good solution, as you consider different possible options:
- Other puzzles (if you do them right)
- Other Black Doors
- Very obvious traps
- Guardians with Riddles
- Other magic items (maybe)
- A cylindrical tunnel, lined with frictionless material, sloping upwards
- A 20' chasm with no bridge
Some things that I believe do not belong in this group:
- People that must be spoken to and negotiated with, or killed, avoided, etc.
- An ordinary locked door, because everyone already knows the main options to deal with ordinary locked doors and thus require no thought. At least: the first locked door may be part of the prior category, but the 50th one will not be. Same logic applies for the 20' chasm if you see a lot of chasms.
You don't have to know what is behind these obstacles for them to work. The impediment may be metaphorical rather than physical (if the only reward is knowledge). I think you could potentially have something like this with nothing behind it at all, if you wish to either teach your players how something works (and equip them better for a similar obstacle) or simply waste everyone's time.
Unique "correct" solutions are not necessary, and in extreme cases no intended solution need be prepared at all (as long as the GM is reasonably confident that the party could come up with a way to proceed).
There also doesn't have to be an explicit penalty for doing the "wrong" thing - it's pretty much the defining quality of a trap, some puzzles use it, the only cost may be time, or it could be entirely negligible if you and your players are happy just solving puzzles. Still, I think that I would recommend incorporating some cost, no matter how small, in general.
Update:
I'm going to refer to these as Special Obstacles until further notice. That differentiates them from creatures and the most common structural features and describes their function in the universe.
Alternative names included simply "Obstacle", "Hazard", "Toy", "Diversion", "Thingy", and "Jawn".
I had another discussion about the folly of trying to categorise things and whether this is fundamentally limiting, but I don't know what to think about that yet. I think my main purpose in digging into this concept is not to try to rigorously and rigidly define the pieces that I am using to stock my adventures; it is to analyse how all these different kinds of things may be presented best and most freely when encountered. So that I can see the similarities between the game structures in different places and save myself mental work.
I also realise that the role that anything will play in your game cannot be described purely by its own qualities and place in the universe. That has a major impact, but there is also the part of how the Party relates to it, which is based on what they think of it, which the GM cannot control. Things can even be deliberately misrepresented, or miscommunication can occur - for example, the story of The Gazebo is famous. Things won't be boxed in.