5 Thoughts on Mystery Planning

In one of the bigger RPG forums in Israel, there's a discussion about investigative games: Possible, or not? And if possible, how can it be achieved? As a person who GMs quite a lot of investigative games, I started to think about how one should plan a game like that and how it should be GMed. My thoughts about the subject follow:

1. The most important thing to consider is what place clues should take in your mystery. In some investigative books, the investigators like Sherlock or Poirot, finding the clues is not the challenge, and both Sherlock and Poirot find them quite easily. In other stories, we can go for quite a lot of time before a single clue is found. For me, I think that the real challenge shouldn't be to find the clues, but to use them together to find the real criminal.
2. Red herrings. This is another important thing to consider, as this concept is used in almost the entire investigative canon of books. For me, I always thought that red herrings are one of the more unique qualities of the genre, and one of the hardest to pass to RPGs. While in other mediums, you can decide when the investigator will throw the red herring, in RPGs you cannot. If I want to use it (and boy, I sure want), I tend to use it as a comeback in a later investigation, or more likely as a small level. You solved the mystery, but there's still a hidden layer that you might need to find, someone who activated the criminal that you caught.
3. Types of crimes. We have to face it, solving mysteries can become as tedious as dungeon crawling (and far more frustrating, of course). Because of that, we have to rewind the mysteries. One crime will be a murder mystery; another will be a kidnapping mystery... But we can go to even farer places: One mystery will contain a moral dilemma; another will confront one of the characters with his/her deepest fears...
4. Complexity levels. I've written about it once, but it worth repeating: Not all mysteries were born the same, some are harder than the others, some are more complex than the others, and some have a bigger scope than the others.
5. Team work. In the canon, the investigator never was alone. Sure, s/he did almost all of the work, but there was another one. In RPGs, we can't have one player who will be the mastermind and the others to be the sidekicks. One way to solve it is with having one person to play the computer genius of the group, another to play the interrogator of the group and so on... A little better approach (to my belief) is to have this allocation of experts, but to have skills that are known to more than one expert. The computer expert might be better at using the Internet, but she will also be good with sneaking past cops. The sneaky might be better at sneaking past cops, but he'll also be good with using the computer. This way, there's a greater feeling of teamwork, as the characters can help each other...

So, these are my thoughts on the subject. How about you? How do you plan/GM investigative games?