Saturday, 22 September 2012

Trappery, part five

In the last issue of Trappery, which was longer ago than I intended, I started thinking about “real-world” factors that affect trap design. As promised, I’ll now have a go at applying these to devise actual traps.

As a reminder, our core questions are:

  • What’s the main objective of the trap?
  • Who or what are the targets of the trap?
  • Who uses the area, and who has authority to bypass or deactivate the trap?
  • How serious is the threat you’re guarding against?
  • How dangerous is your everyday life?
  • Are you military, official, commercial, criminal or a private citizen?
  • What sort of budget and resources are available?
  • How easily can you maintain or inspect the trap?

One other thing to bear in mind. Realistically, you should never be designing “a trap”. Proper security relies on a combination of factors, which may or may not include traps, but in any case work in tandem. In some cases, a trap may be the best solution to a particular security problem, but it’s only part of the system.

Note: Originally I began this by showcasing some sample situations to consider trap options in, as the previous comment suggests but they ended up really quite long even before I'd looked at any specifics. I decided to swap those out and post them separately.

The Tomb of the Iron Queen

The corridor

It's sort of like this

Here’s our example. We’ve got the classic tomb-raiding scenario going, and today’s trap is designed to protect a corridor leading to the weapons chamber of the tomb. Let’s assume it’s a non-fantasy setting for now.

  • Objective: stop people getting to the weapons chamber, or escaping with any loot if they succeed
  • Targets: anyone who enters the corridor, but not animals and the like
  • Clearance: nobody should be traversing the corridor, and nobody’s expected to maintain the traps
  • Threat: we’re trying to protect a tomb in eternity, and they have a tendency to get robbed. The deceased object very strongly to that kind of thing. Because there’s nobody to monitor the tomb in the long term, traps need to be pretty final, though deterrence is useful too.
  • Lifestyle: it’s an old tomb from more savage times when life was cheap, especially for powerful queens.
  • Standing: this is probably an official trap, designed to repel robbers rather than armies
  • Resources and maintenance: the queen has all the money and talent she wants for the initial work, but there’s no real way for anyone to maintain traps, monitor them, refill or refuel them or otherwise keep things running.

Okay. We know that nobody at all is supposed to be in this corridor. We know the queen has little regard for life. We know that there’s nobody to retrieve prisoners or respond to alarms. We know there’s a strong probability of intrusion and the builders consider that a serious issue. What that means is that the corridor is going to feature traps that kill people. Not capture, not inconvenience, not even maim, but kill.

We also know there’s nobody to maintain the traps, and that we want them to function in the long term. Um, let’s ignore the fact that realistically pretty much any trap would stop working after a few decades without maintenance... it’s a game. But this does mean that we have to consider reusability and lifespan. A trap that’s destroyed in triggering isn’t that sustainable, nor is something with only one shot. We should also bear in mind that most poisons lose potency quite fast, so they won’t be as much of a feature as we might like. Animals are also not a great bet. On the plus side, tombs tend to be dry, so less decay to worry about. Things you can generally rely on in the long run include gravity, force and other laws of physics.

For targeting humans, you want to discriminate in a fairly broad way. Lacking magic, high technology or anything with decision-making powers, you can do this on size, action (who else opens locks?), or weight (versus, say, rats). Things like tripwires and other delicate mechanisms aren’t a great bet, either because of decay over time, or because they’re too easy to trigger.

Okay, at this point I need to make a semi-arbitrary decision about trap type. As I’m protecting a corridor, I’m going to go with the classic trigger-stone trigger mechanism. A weight suggestive of a human foot will set chains and counterweights in motion within the walls, ultimately causing spikes to erupt from the walls in a ten-foot stretch. Stepping into the trigger area (towards the end of the six feet) lets counterweights drop and force out the spikes. Spikes would be pretty fatal against lightly-clad tomb raiders and the area’s big enough to catch a pair of them, while small enough to keep the mechanisms relatively simple.

So for this to be a reusable mechanism, the counterweights need to be raised between uses. How can we do this? One option is a fluid recharge, where the counterweights are cranked up by the flow of an underground stream or similar; you could use a reservoir of water (or sand, or liquid mercury) that drips out when the trap’s been triggered and builds up enough weight to get it primed again, but that will soon run out unless it’s truly massive. Alternatively and fiendishly, you set up another mechanism elsewhere in the tomb that causes the raiders themselves to prime the trap. This might use the pressure of their feet on corridor, or the force of them opening a door, or any other method of capturing force, to raise the weights and prime the trap. I quite like the idea of having the doors at the entrance end of the corridor be wired into the mechanism, so opening them actually arms the spikes. Let’s go with that, ignoring whether or not you could do that in practice.

Practical Tomb-Robbing

From an adventurer’s point of view, here’s how the trap looks:

You find a pair of heavy, ornate doors bearing carvings of soldiers with all manner of impressive and exotic weaponry.

A successful [insert appropriate skill here] roll suggests that this is an armoury chamber, where the queen’s favoured weaponry has probably been interred. Most likely there’s also an array of extra weaponry as death-offerings to make the place suitably grand.

There’s no sign of any traps on the doors; they aren’t even locked. The doors are heavy and take some effort to open. A successful Listen roll reveals the faint sound of moving chains within the walls. An Engineering roll suggests the doors are counterweighted in some way, though a good roll suggests the doors could be linked to some kind of mechanism elsewhere. You see a classic ten-foot corridor stretching into the darkness. The walls are heavily carved and decorated with geometric patterns and animal, whose blank eyes seem to stare at you. Examining the eyes reveals they are actually holes in the walls. Roll Engineering or History, and also Dungeoneering. Engineering or History suggests several possibilities: this might have allowed air circulation during construction, or they might have had gemstone eyes that have been stolen, or eyes of some material that’s since rotted away. Dungeoneering points out that these would also be handy for spying on intruders (but who is there to do that?), pumping in poisons, flooding the chamber, or firing darts. The floor is made of red and yellow tiles about a foot square. Overhead, there is a vaulted ceiling whose painted surface has long since faded and peeled away.

The corridor stretches on for some distance, narrowing to about five feet in width. If you’re paying attention to the carvings, you’ll notice the animals become more frequent as the corridor continues. One stretch seems to feature a very large number of owls. If you’re not paying attention, the DM will call for a Perception roll, which may draw your attention to the owls. A History or Religion roll can reveal that these were considered vengeful guardian spirits of the dead (which should raise a little caution). If the Perception roll was very good, or if you examine the walls here, you will notice that this means the wall here is full of holes for a stretch of about ten feet. They’re not a regular pattern, they fall naturally wherever the owls happen to be, but they are very frequent here. Depending how much information you picked up earlier, a kindly DM could call for Dungeoneering here either instead of or as well as the Perception roll, to highlight that this might indicate a trap. If you try to assess them as a trap, you can easily deduce that anyone standing anywhere in this area would be covered by between three and ten holes at all times.

The holes are deep and hard to examine visually (and that might not be very sensible). If you try poking something inside, you can evoke the faint clink of something on metal. The metal objects seem to be found in every hole here. Comparison with holes earlier on reveals that those seem to be empty.

If anyone ventures into the last five feet of the owl section, and the weight they exert on the floor is more than a few pounds, the tiles suddenly give under their feet. There’s a frantic rattle of chains and three-foot metal spikes burst from the owls’ eye sockets, burying themselves in anyone unfortunate enough to be in front of them. Not only will they be injured badly, but it’ll prove difficult to reach them to administer help, and for them to get past the spikes. They’re tied into some heavy mechanism and though foot-wide sections can be forced back with an effort (Strength check), they slide back if not prevented in some way.

Once the trap has been triggered, there are a couple of options to bypass it. You could force the spikes back and wedge them somehow. You could actually clamber over the spikes, since the vaulted ceiling gives room overhead. You could also squirm underneath, since they don’t go all the way to the ground.

If you identify the trap before it triggers, you could try to avoid triggering it. If you find some way to exert a few pounds-worth of force on the trigger zone without actually venturing inside (and bearing in mind that it only starts five feet into the owl section) you’ll set it off without the pain. Blocking the holes will be time-consuming and call for a roll when the trap triggers, since it might simply force out any blocks. You could squirm underneath (grubby and slow work) so that even when it triggers you’re basically safe, though you’ll have to drag anything like backpacks behind you and the trigger moment might be pretty tense. You could wedge something into the holes either side of the wall to create a crude scaffold and climb over without touching the floor (requiring a crafting roll, and possibly a Dex roll depending on the scaffold they build).

Alternatively, the enterprising adventurer might want to disable this trap. This is where you roll Disable Device, right?


Disabling Devices

A little while ago, Dan H over at Dreamers and Dicepools made one of the most insightful remarks I’ve heard on RPG design. It’s the direction I already wanted to go with traps, but I think it sums up the issue in a more compact and transferable way than anything I could have said.

Dan was working on a stealth-based P&P RPG, where you play sneaky assassins, and had already noted that stealth gameplay is very poorly supported in most RPGS. How to implement stealth in his own game, he pondered?

The first clear design principle that emerges, then, is this: under no circumstances should this game have any kind of Stealth skill.

D&D doesn't have a skill for fighting, or for exploring. Vampire doesn't have a skill for using vampire powers. Paranoia doesn't have a skill for accusing people of being traitors. The core elements of gameplay are always either modelled in great detail (D&D devotes pages to combat, every White Wolf game devotes pages to supernatural powers) or entirely absent (the Warhammer 40,000 RPGs don't have any rules for being devoted to the Emperor, but the Imperial Creed permeates the setting anyway).

I think the point I'm groping towards here is that Stealth skills make stealth a purely abstract concept. Want to sneak past the guard? Make a Stealth roll. Want to hide from your target? Make a Stealth roll. I don't want to sound like one of those idiots who talks about using "active verbs" but the traditional RPG Stealth roll puts the emphasis on the consequences of the player's action (hiding from the enemy) rather than process by which they achieve it (ducking into a wardrobe, jumping into the rafters, hiding behind a curtain). It's the equivalent of D&D combat being resolved by a single "kill all the monsters" roll.

If a game is going to have Stealth gameplay, stealth is going to have to actually be gameplay, not just "roll move silently to sneak up on the Orcs." And gameplay is all about choices and interactions.

While I haven’t actually expressed this idea, what I’ve basically been trying to do with Trappery is to change traps from being fringe elements of gameplay to core ones that are interesting to interact with. As Dan points out, resolving elements of play through a single roll makes those elements peripheral. Shannon at ST Wild has put it even more bluntly with a brief article called “Roll X to Win”, which just about sums it up.

Jump is a fairly sensible skill because you just want to find out whether you make it over the chasm or up to the branch, and that uncertainty can make for some thrilling moments, but jumping per se is not a fascinating part of gaming for most people. Finding a way past a murderous deathtrap ought to be a little richer. The Disable Device skill, as applied to traps, is actually one of the main obstacles to this. We want traps to be gameplay.

Back to the Tomb

There are several options for interfering with this trap. We’ve established that it relies on a system of chains and counterweights in the walls. You could break into the walls to get access to them and either fasten or break the chains. Any rolls required here would basically determine the effort and time required rather than ultimate success, since there’s no real failure chance, unless they completely botch the wall-breaking and start collapsing the tunnel... an Engineering roll might be needed to do that safely. Attacking the door end of things might be safer, since the chains presumably attach to the door somehow, and demolishing the door is less likely to bring the roof down on you. You could trigger the trap and then painstakingly bend the spikes, so the trap is both ineffective and unable to reset. You could carefully excavate the floor and tamper with the triggering mechanism itself, though that’s more dangerous if the trap’s still active; here I’d use some kind of Dexterity and Strength rolls to do the digging, and Disable Device to interpret and then manipulate the triggers correctly. You could bring in a load of half-inch steel plate and attach it to the walls so the spikes can’t emerge. You could attack the owl section with sledgehammers, smashing it apart until you have access to the spikes themselves and then taking the spike plates off the chains; again, this creates some structural problems.

I didn’t highlight this at the time, but there are also several ways of spotting the trap exists and what it is. Obviously, there’s metagame player suspicion that a tomb will be trapped. There’s also reasonable in-character suspicion. The effort and noises involved in opening the door are a clue that something may be up. The holes in the wall should attract attention, and if not various skill rolls can flag up the danger. The precise nature of the danger is fairly easy to establish at that point, and the danger zone. So the issue becomes less a matter of “will they find, escape and disable this trap?” and more a case of “how will they find, escape and disable this trap?”.

Of course, having it this simple to detect the trap means (from the builders’ point of view) it’s possible that would-be thieves would realise there’s a trap rather than walking into it, but providing they can’t bypass it deterrence is an acceptable result. At the time of building, they probably wouldn’t expect a band of humble tomb robbers to have the skills or resources needed to get safely past the trap. So I don’t consider that a big problem.

So that's it for this edition of Trappery. Next time I'll probably post up my security planning examples, which touch on the way that traps are part of a coherent strategy rather than random isolated things. Any requests for future discussions?


  1. Something that I think this example illustrates very well is that thinking through one element of desgin can make the whole experience more interesting. Working out how a trap is supposed to be reset, for example, means working out how it connects to the rest of the dungeon, which in turn allows you to have a clearer idea of how the players might actually detect and disarm the trap.

    There's sort of a vicious/virtuous cycle here. The more abstracted you make a trap, the more you have to treat it as a purely game mechanical challenge - if you don't really know how a trap is triggered, then you have to rely on Spot/Disable Device rolls to find and deactivate it, which makes the whole thing very hit-point-tax-ey. Conversely if you start to think about how a trap actually *works* you can start thinking about ways you can actually *avoid* it.

    1. That's a very interesting point. I'm inclined to say part of the difficulty is the different levels of abstraction between traps and combat. If you're designing an Evil Dungeon in a hurry, you can grab a few monsters that're about the right level to guard a room. When the PCs encounter them, there's a well-developed combat system to handle a range of things that'll come up in a fight. There's also at least some mechanical support for non-combat alternatives. It's not going to be the most immersive encounter ever if you haven't thought about their relationship with other elements of the Dungeon, but there's enough mechanical framework that things'll probably work out okay. If you try the same thing with a textbook trap, they're so abstracted that there's not much to work with and player options seem a lot more limited.

      Combat is elaborate enough to allow huge scope for players to do cool and inventive things in the course of winning a battle. It could easily be the other way.

      "Roll Perception."
      "I fail."
      "Okay, you're ambushed by a band of goblins. Roll Fight."
      "I pass."
      "You take 8 chewing damage. Do you want to loot the bodies?"

      "Roll Perception."
      "I pass."
      "You hear some goblins approaching."
      "Well, we do need to get to the stairs... I'll attack them."
      "Roll Fight."
      "I fail."
      "You take 23 chewing damage. Do you want to loot the bodies?"

      Sounds like fun, right?