Aug 18 2009
How to Foil Investigation Magic
When people start talking about creating mysteries in settings with magic, the magic often comes in as a potential problem. How do you hide the murderer when a simple spell will expose the evil in his heart? What do you do when dead men can tell tales? Is there still a mystery if someone can use magic to see exactly what happened at the time and place of the crime? In short, is investigative magic going to break the mystery?
The answer to this last question is no, not necessarily. It might make it a bit more difficult, but it doesn’t have to break it entirely—and it’s possible to keep investigative magic from being a game-breaker without saying “No, it just doesn’t work.” If you’ve read my post on bypassing lie detection, you’ve already got part of the idea; it’s all about knowing exactly what the magic can and can’t do, and working around that.
Let’s start with an easy one: D&D’s detect evil, particularly in the third editions where paladins could use it pretty much constantly. At first glance, it seems like trouble, right? But all it can do is say that somebody is evil. If you’ve got several obviously nonconnected evil people in the town, it might narrow down the suspect list, but it’s not going to immediately give you a culprit. And that presupposes that the culprit isn’t using some sort of anti-detection magic—and, for that matter, presupposes that the culprit is actually evil. Moral ambiguity has its uses; particularly if you’re not dealing with a murder, who says whoever committed your crime has to be evil? In short, use the rest of the people around as protective camouflage.
Here’s a slightly tougher one: what do you do if you’re dealing with someone who can ask the local animals what they saw? Unless they know there’s a detective with that kind of MO out there, most criminals aren’t going to try to eliminate them as witnesses or take preventative action. But that doesn’t mean it’s going to give them everything, either. Most animals have their own limitations when it comes to giving descriptions people would find useful: not seeing in the same spectrum, not understanding which facts are important and which aren’t (and thus leaving out some of the vital ones), short or selective memories, you name it. In short, they’re unreliable witnesses. Just take advantage of that.
One of the most potentially troublesome pieces of magic I’ve seen allows its caster to actually recreate the scene of the crime if she knows where and when. This one’s harder to work around, but not impossible. Assuming that it can’t be blocked entirely, most spells of that sort that I’ve seen are purely visual, or occasionally visual and auditory. If your culprit knows that such spells exist, or even just wants to take into account the possibility of being seen, she might utilize disguises—to look like not-her, or even to look like someone else if she’s feeling ambitious.
What about the dead? Many murders in our world come about because their perpetrators don’t want to leave living witnesses. But some magic allows for talking to the dead, some people leave ghosts, and in some worlds it’s even possible to resurrect the dead—and all of those give you one material witness with a bone to pick. A killer who doesn’t want to leave traces will need to know how to bypass such tricks. Some work by destroying or trapping the soul, leaving nothing to provide answers. Others destroy or mutilate the body so that it will be beyond the effects of either revival or communication. Still others don’t bother with that; instead, they just make sure their victim either doesn’t see them or gets the wrong impression of who did it.
In sum, investigation magic doesn’t have to break your plot. It just means you need to think.
I think you’re on the right track with these ideas, however I think instead of dumbing down the results (Animals and even the dead people beeing unreliable witnesses), it’s more satisfying for the players if the magic they mastered helps (The dog they question remembers that the Murderer smells like all kinds of fresh meat giving a strong hint at the local butcher) or even better, they actually NEED it to solve the case.
Magic’s only an additional tool, and it’s a tool for the murderer as well as the investigator:
1) A geas or mind control is difficult to prove even from a magical scene recreation. The fact is, even if we just KNOW in our hearts that Sally isn’t the murderer, if she’s right there going stab it’s hard to argue with the “infallible” magic.
2) Endemic problem creatures, like the undead, can often be created and commanded by people. Unless there is a way of telling the difference between purpose-built undead and residual undead, there’s no way of proving their use as a murder tool. (Bob was ‘et by zombies. Turrible sad that.)
3) Some magical effects may be so natural-looking (or so outre) that no one actually believes that they are murder - i.e. lightning during a lightning storm, or alternately telekinetic heart fibrillation.
4) If people don’t trust magic implicitly then any special knowledge of the murder might make the mage a suspect…
Which raises an additional point. In early modern times, poisoning was considered a much worse crime than, say, stabbing. The stabber was hanged, the poisoner was boiled to death.
Is using magic to kill people similarly “frowned upon” in fantasy societies? (Or evil miracles for that matter?)
I wasn’t so much trying to make it useless as to keep it from being a win-button; I know a number of people have tried to nerf effects I’ve used because they couldn’t figure out how to provide help without necessarily providing a solution. If you’ll notice, having unreliable or incomplete witness creatures gives you information without giving you everything. The detect evil trick either narrows down your suspect pool rather than finishing it, or implies that things are rather more complicated than expected. The speak with dead one–for a while, I ran in places where “How does speak with dead not ruin my mystery?” came up every other month or so, so getting it out of the way seemed important.
And you might be interested in the post to which this was supposed to be a precursor, basically a riff on how even the absence of evidence can be evidence. It’s the problem with a daily update sequence and the inability to edit after a certain duration; the writing order sometimes ends up suboptimal so the links will still work.