Tuesday, September 11, 2012

Investigation and Evidence

When I run a procedural or investigative game - and I do love to run them, in all kinds of different genres - I never have to worry about the PCs finding the right clues. No, I don't automatically make clues obvious, nor do I scatter ten clues in the hopes they'd find one. Not at all.

Then what DO you do, clash? Pray tell!

I don't use 'clues' at all.

What? Heresy! How can you investigate without clues? 

The only baddies that leave clues behind are the ones that are framing someone else, or leading the investigators on in some other way.

Huh? WTF are you babbling about clash? Is this one of your quixotic windmill tilts like 'illustration' vs 'art'?

No, I don't think so. I think it's a very important conceptual step in creating a fun and interesting investigative game.

Well, go on with it, then. How do you investigate without clues?

You investigate by looking at the evidence. From this evidence, the investigators extract the truth. The GM should not 'place' clues, but rather should let the players extract information from the body of evidence. The players establish what is and what isn't a clue, not the GM.

Ummm, how's that work?

OK, example time. Let's take a murder investigation and run with it. So, the GM knows who committed the crime, when the crime was committed, where the crime was committed, how the crime was committed, and why the crime was committed. From this the GM describes the murder scene as a general gestalt when the PCs first arrive to examine the scene. The players now begin their investigation.

The PCs use their various skills and abilities to extract detail from the evidence, first at the murder scene - fingerprints, wound patterns, skin lividity, body temperature, insect larvae, blood patterns, bullet trajectories, etc.; and gradually elsewhere - phone logs, financial records, witness interviews, surveillance tapes, and the like.

Let's call this the body of evidence.

From this body of evidence, the players - and their PCs - use logic and intuition to see patterns in that evidence, and from these patterns, begin to reconstruct the crime. They develop a list of suspects. They do more investigating, focusing on these suspects and this possible reconstruction. It's an iterative approach - focus, investigate, find patterns, refocus, etc. - which winnows out suspects, means, and motives until only one or two solutions fit the established facts.

Now some of these facts may, in fact, be suspect, and some may be unsuspected lies. Witnesses may lie to protect themselves. Alibis can be faked. Even clinical evidence can be deliberately misleading - the right-handed killer may have struck with his left hand in order to muddy the trail - this would be an example of a false 'clue' planted to throw off investigators. Things is, if nothing fits the facts, the facts must be questioned and the evidence re-evaluated.

There really is only one possible complete solution - the truth - but depending on what patterns the investigation sees (no investigation uncovers all the facts) there may be a very small number of possible *partial* solutions which *do* fit the known facts. These partial solutions have to be concentrated on and incorrect fits eliminated until only one fits.

That sounds like a lot of work, clash. Might as well be a detective.

Well, if you are going to model something, model it correctly. That's my motto! This way, the player will feel like detectives. :D



  1. I don't understand the distinction you're drawing between 'evidence' and 'clue'.

  2. Hi John:

    There is a style of GMing that has been around for a long time, where the GM crafts discrete 'clues' to be found by the players which, when assembled together, point in the desired direction. This can lead to the GM leading the players around by the nose, or to frustration as the players miss these carefully crafted clues.

    I think a large part of this lies in a common connotation of the word "clue". A clue is what you ask for when you can't figure out a riddle or puzzle. "I dunno Brian, I don't even know where to start. Can you give me a clue?" In this sense, the clue is discrete - not intrinsic to the puzzle, but supplied extrinsically. Evidence is always intrinsic to the problem, and it is the pattern of facts given by the evidence which leads to the solution.