How to Solve a Murder Mystery: A Beginner Guide to Deduction

Most people who are new to mystery games do the same thing: they interview every suspect once, politely, and then guess whoever seemed shiftiest. It almost never works, because seeming shifty is not evidence. In a well-built mystery, the innocent suspects are all hiding something, and behaving guiltily is exactly what a person does when they have an affair, a debt or a theft to conceal.

Here is a method that works better. It is not complicated, but it does require doing things in the right order.

Start with the weapon, not the killer

Whodunit is the interesting question and the wrong one to start with. Begin with the physical facts, because they cannot lie to you.

What killed this person? Where were they when it happened? Does the weapon require strength, or reach, or specialist knowledge? Was it brought to the scene or picked up there? A weapon taken from a locked study tells you the murder was planned by someone with access. A weapon grabbed off a mantelpiece tells you it probably was not.

Every answer here quietly eliminates people, and elimination is more reliable than suspicion.

The three-legged stool

Means, motive and opportunity. It is a cliché because it is correct, but the useful insight is about which leg is scarcest.

Motive is usually the most abundant. In a good mystery, most of the cast wanted the victim gone. Motive alone narrows almost nothing and is where beginners spend most of their time.

Means is narrower. Not everyone could have used that weapon in that way.

Opportunity is usually narrowest of all, because it depends on a timeline that other people can corroborate or destroy. Spend your effort here.

Build a timeline before you build a theory

Write down every claim about where someone was and when, as they make it. Not your interpretation of it, the actual claim.

A timeline does two things at once. It exposes gaps: the fifteen minutes nobody can account for is where the murder lives. And it exposes collisions: two people cannot both have been alone in the library at nine o'clock, so one of them is lying, and you now have a specific, provable thing to press on.

Do this on paper or in a notes app. Holding a timeline in your head is how people convince themselves of things that are not true.

Ask closed questions before open ones

The instinct is to ask "tell me about your evening", and the answer is a rehearsed paragraph. Rehearsed paragraphs are useless.

Closed, specific questions are much harder to deflect. Where exactly were you at nine? Who saw you? What were you wearing? Did you hear the clock? A liar has prepared a story, not a world, and detail is where invented worlds run out.

Then, and only then, go open. Ask what they thought of the victim. Ask what they think happened. Once someone has committed to specifics, their loose talk becomes checkable.

The contradiction is the whole game

Suspicion moves nothing. A contradiction moves everything.

In BACKSTABBERS, this is a formal mechanic rather than a vibe. When you hold a piece of evidence that genuinely conflicts with what someone has told you, you can present it to them, and if the conflict is real they crack. If it is not real, nothing happens except that you have shown a possible murderer what you have got.

So the sequence is: get them to commit to a specific claim, then find the thing that disproves it, then confront them. Confronting first, on a hunch, is the most common way to waste a run.

Secrets are not guilt

This is the distinction that separates good players from lucky ones. In our cases, the murderer lies. The innocent suspects tell the truth — but every one of them has a secret they would rather not surrender, so they dodge, minimise and change the subject.

That means evasiveness is not a signal. Everyone is evasive. What matters is whether their account, once you finally pin it down, is false. Someone squirming about why they were in the corridor may be hiding an affair. Someone whose stated location is contradicted by physical evidence is a different matter entirely.

Our piece on alibis goes further into how a fabricated account differs in texture from an embarrassed one.

Beware your first suspect

Once you have a favourite, you will unconsciously read every subsequent answer as confirming them. This is the most reliable way to lose a case you could have solved.

Two defences. First, force yourself to name a second-best suspect and argue their case honestly before you accuse anyone. Second, ask yourself what you would need to find to prove yourself wrong, and then go and look for it. If you cannot describe what would change your mind, you are not investigating any more; you are decorating a conclusion.

Practise on something that resets

Deduction is a skill and it improves with volume. The awkward part is that most mysteries can only be solved once.

The Daily Case exists partly for this. A new case every day, about twelve minutes, no account required, and the same case as every other player in the world — so you can compare your route through it with somebody else's. Twelve minutes a day for a fortnight will do more for your instincts than reading another guide.

When you want to work at proper length, the full-length investigations give you up to an hour at Rookie difficulty, with three guesses to spend. If you would like the pressure turned up, the difficulty modes go as far as twenty-five minutes and one guess, which punishes exactly the sloppy habits described above.

Facts first. Theories later. Accusations last.