I don't mind traps being a focus for rogue players, so long as everyone else gets something that's also focussed on them. I also tend to think about traps in at least two different categories: combat and exploration. I do have a couple of traps in the interaction category, too, but those are still relevant and there's a small chance one of my players would find them if they're mentioned here
For traps in combat, I go with simple effects that are easy to understand (at least, simple the second time): pressure plates that cause darts to shoot, pits, etc. In one fun combat, I had a room with chessboard tiles where every time you stepped on the white tiles, a valve opened above and dumped just enough water in the room to fill it an inch deep in water. This caused the party a little confusion, since they didn't see the threat.
Then the doors locked and the air elementals attacked and started pushing them around. That was fun.
For exploration traps, I favour traps where having a rogue is helpful but not essential and, once the trap is found, anyone can do
something to attempt to deal with it. The key design feature to me is that every trap have a purpose. This is important for combat traps, too, but especially so for exploration ones because they will often be found on their own. Some purposes I've used before include (translated to 5e terms):
- Kill anything of the expected challenge rating that attempts to access the forbidden area
The expected challenge rating is important. Kobolds might expect to face anything up to an equivalent of a CR4 party - they don't have the resources to make a trap that's very likely to kill someone stronger, but their primary goal is to kill anyone who stumbles upon them.
- Do as much damage as possible and summon the nearby guards
Of course, the nearby guards may not have been present for centuries, but the trap's original purpose doesn't change.
There's a strong distinction between traps that are designed to kill and ones designed to injure. Certainly the builders of the latter kind would generally be happy if the threat is killed by the trap, but it's clear that relying only on traps is a risky business, while having traps and guards is much more likely to keep something safe if you expect to have the resources for it.
- Sound an alarm without being noticed
- Scare people away
The most complex kind, usually.
For traps like this, I usually use something from the "kill anything" school, but go for mechanisms that are obvious, defended against tampering and tend towards overkill. For instance, an obvious area in a narrow ravine where a square on the ground has been cleared. Light objects hit a weak magical shield that forces them to pile up around the edges of the square. Heavier objects cause the square to depress about an inch into the ground and barbed spikes to rush out of obvious holes that run in three columns all the way up to the top of the ravine. The spikes are offset a little to each side, and at full extent they reach about 60% across the gap (so anyone caught in the trap is going to be probably take enough damage to be dead in one hit). After a few seconds, the square plate moves forcibly upwards and the spikes retract. If anything manages to block the plate from moving upwards, a wand of disintegrate is triggered. If the mechanism detects that it has been unable to work for a full minute, three wands of Mending are triggered at strategic points.
This trap is extremely obvious, and of course there's a relatively easy DEX save to jump back when the square plate sinks. The point is to find ways around it.
I've tried it on three parties - no deaths, but a lot of fun roleplaying as the whole party tried things out. One of them even managed to break into it and recover the wands.
