A case file is not a story told in order. It is a pile of papers written by people who each saw a small part of what happened, and some of them are wrong. Your job is to put the pile in order, find where the papers disagree, and decide which one to believe.
This is the method we use when we test our own cases. It works for any case-file game. The examples below are invented for this note; none of them comes from a Murder Desk case.
1. Start with the briefing, then read everything once
Most new players skip the briefing and go straight into the files. Read it first. It sums up what the original investigation decided had happened, and it lists exactly what you have to work out.
Then read the rest of the file once, without taking notes. On the first pass, read fast. Do not try to solve anything yet. You only want the shape of the case: who the people are, where the places are, and roughly when things happened.
Notes taken on the first pass tend to fix your first impression in ink. Wait until you know what is in the file before you decide what matters.
2. Build the timeline before you build a theory
Go through the file again, this time with a pen. Every paper that carries a time goes on the timeline: receipts, tickets, logs, phone records, letters with a date, and the time a witness says they heard something.
Write three things for each line:
- the time;
- what happened;
- where you read it.
The third column is the important one. “19:37, coffee paid for at the station café (the receipt)” is evidence. “About nine, he left (his sister says)” is a claim. Both go on the timeline, but they are not worth the same.
When the timeline is done, look at the gaps. A gap of forty minutes in the evening of the death, for the one person who says they were at home, is where your questions start.
3. Keep what a witness saw apart from what they decided
People rarely lie as much as they assume. Read each statement and mark every sentence as one of these:
- seen or heard: “I heard a door slam just after the news started”;
- decided: “so he must have gone out”;
- told: “my neighbour said she saw a light”.
Only the first kind is evidence. The second is the witness doing your job for you, and often doing it badly. A witness who heard a door may be right about the door and wrong about whose door it was.
4. Test every account against paper
An alibi is a claim. A ticket, a till receipt, a timesheet or a logbook entry is a record. When a claim and a record disagree, the record usually wins.
Usually, not always. Ask three questions about every record:
- Who wrote it, and did they have a reason to get it wrong?
- When was it written: at the time, or afterwards from memory?
- Does it prove what it seems to prove? A receipt shows that something was paid for. It does not show who was holding the wallet.
5. Look for the paper that should exist
Missing evidence is evidence too. If someone says they took the last train, there should be a ticket, a guard or a timetable that agrees. If there is nothing, write that down.
Look the other way as well: a paper that should not exist at all. Why would anyone keep that note, or write that letter on that day?
6. Keep every person in play until something rules them out
It is easy to like or dislike the people in a file. Neither is a reason. Keep a short list of everyone who could have been there, and cross a name off only when a paper puts that person somewhere else. Write the reason next to the line.
A suspect you crossed off on the first evening because they seemed kind is a suspect you never really tested.
7. Opportunity, then means, then motive
Work in this order:
- Opportunity comes from the timeline. Who could have been there?
- Means comes from the scene and the reports. Who could have done it that way?
- Motive comes last. It is the easiest part to imagine and the hardest to prove, so do not let it lead.
8. Before you vote, prove it with two papers
On Murder Desk, a verdict has five parts: who did it, how, why, the period when that person’s own account is false, and the two exhibits that prove it best. The last part is the test of your whole case. If you cannot point to two papers that together show it, you have a strong feeling, not a solution.
Go back to the timeline, find the moment where one account and one record cannot both be true, and name both.
9. Ask for help early, not late
Every case has hints from Sherlock Holmes. They start vague and get more precise, and each one costs points. If you have been stuck on the same question for twenty minutes, take the first hint. It usually points you back to a paper you have already read, which is the best kind of hint.
Try it on a short file
The training case, Closing Time, is short enough to try this method in one sitting. Open it, build the timeline first, and see how long it takes before one of the accounts stops fitting.