What Makes a Good Detective Game?

A good detective story makes the work look irresistible. In Mystic River, the murder of Katie Markum is solved through interviews, leads, evidence, and careful attention to the facts. The detectives gather details from the crime scene, question people close to the victim, notice what does not add up, and eventually start seeing the shape of the case.

That is exactly the fantasy detective games often promise. The player should be able to turn clues and witness statements around in their head until the all-important eureka moment arrives. The strange thing is that many detective games do not quite produce that feeling, even though the genre has existed for decades. Classic point-and-click adventures, side quests in open-world games, and full detective adventures have all tried to capture it, but the player rarely feels fully responsible for cracking the case.

The central design question is therefore simple: how can a game let players unleash their inner Sherlock? To answer that, it helps to separate the job of detective work into several pieces: gathering information, exposing lies, following leads, connecting evidence, making deductions, and finally proving an accusation.

The first step is gathering information. Detectives get clues from the crime scene, such as the bullet that killed Katie, phone numbers in her backpack, and a call to the police. They question her parents, her friends, and a witness who saw the crime. This part translates to games fairly cleanly.

LA Noire handles evidence collection with 3D objects the player can pick up, rotate, and inspect. Condemned uses investigation gadgets to reveal fingerprints and bloodstains. Dialogue trees are common across detective games, and they are a natural way to question witnesses and suspects. None of that is the difficult part.

The hard part comes after the player has the information. A detective does not just collect clues. They use clues to expose lies, follow leads, find connections, and make deductions. That is where many games start to struggle, because testing whether the player has figured something out can accidentally tell them what they were supposed to notice.

One common solution is the multiple-choice question. In The Wolf Among Us, Bigby can inspect a clue and then answer a question about what it means. Sherlock Holmes: Crimes and Punishments does something similar during interviews: at specific points, a button appears that lets the player challenge a witness and pick the response that exposes the lie.

The problem is that the player is being heavily prompted by the available answers. They are not fully forming their own thought so much as looking for the option that sounds most plausible. Sometimes they may simply guess. If a wrong answer just asks them to try again, the system becomes even weaker.

This reveals the core challenge. The act of asking the player whether they have solved the puzzle can give away the existence and shape of the answer. Once the game says, in effect, "connect these pieces of evidence now," the player is no longer making the same kind of discovery a detective makes. They are passing a comprehension check.

Some games improve the multiple-choice model by giving the player many more possible answers. Detective Grimoire asks the player to build a thought from several separate pieces: pictures, statements, and slots that can be combined in many ways. The result is still a prompted deduction, but the number of combinations makes random guessing much less useful.

The Trace uses a related approach. The player answers questions by dragging in clues and observations they have found, often needing several pieces for a single answer. Ace Attorney has one of the cleanest versions for exposing lies. A witness gives testimony, and the player presents evidence that contradicts a specific statement. Because there are several statements and a large inventory of evidence and profiles, the player has to think carefully about which contradiction matters. The life bar also punishes misplaced objections, which discourages brute force.

This can produce strong detective moments. In Life is Strange, the player studies an evidence board and chooses the right clues to work out where Nathan Prescott took someone after a party. The large set of possible clues reduces guesswork and encourages the player to look for real connections, such as dates, times, or a broken taillight that identifies a car.

Even here, though, the game is still prompting the player. The question itself tells them that there is something to solve, and often suggests the kind of answer they should look for. The player may feel less like the detective and more like the sidekick being quizzed by someone who already knows the truth.

A stronger approach is to give the player tools for pointing out contradictions and connections without first asking a narrow question. Contradiction does this well. The player interviews suspects, gathers a long list of statements, and can then bump any two statements together to catch someone in a lie.

That is different from Ace Attorney, where the player always knows that a flaw must exist in the current testimony before the scene can move on. In Contradiction, the player can leave one witness, speak to another, come back later, and keep several conversations in mind at once. The player does not know exactly who is lying or which pair of statements contains the problem. They have to follow the case.

Papers, Please is not a detective game, but its discrepancy system shows why this approach is powerful. The player cannot reject someone at the border without providing a reason. To do that, they press the inspection button and highlight two things that contradict one another: perhaps an expired passport and the current date, or a document detail and the rule book.

In both systems, the player is no longer simply responding to the game. They are telling the game that they have spotted something interesting. The Papers, Please version is especially elegant because the player can highlight any two pieces of text on screen. There are too many possible combinations to brute force, and any visible piece of evidence can matter, not only items the game has chosen to put into an inventory.

Contradiction and Papers, Please work well for links and lies, but deductions are harder. Discworld Noir shows both the promise and the limitation. The protagonist automatically writes useful information in a notebook, and the player can combine two notes to form a deduction. In one puzzle, the player connects a strange phrase written on a wall with the fact that the victim was hung upside down.

The puzzle itself is good. But for it to work, the game has to create notebook entries for both the strange phrase and the upside-down body. The second note can give too much away. When thoughts become puzzle pieces, the player is not always using their own logic. They may simply be snapping together the logic the game has already prepared for them.

The Shivah offers a cleaner alternative. The player reads emails belonging to a dead man and notices messages from someone named Ethan G. The question is which surname belongs to Ethan. He might be Ethan Goldwater, mentioned in another email, or Ethan Goldberg, connected to an accounting firm in a ledger. The game does not hand the player those possibilities as inventory items. Instead, the player opens a fictional search engine and types a name. Ethan Goldwater returns nothing. Ethan Goldberg works.

That moment is powerful because the information leaves the game world, enters the player's head or notebook, and then returns through an action the player chooses. The player is not selecting from a list of possible deductions. They are testing a thought they formed themselves.

Her Story pushes this idea further than almost any other game. It gives the player a database of interview clips about the death of Simon Smith, but the only way to see those clips is to type search terms into a computer. Only the first five chronological results are shown for each search.

The player receives almost nothing tangible: no inventory of clues, no formal list of observations, and no notebook full of ready-made deductions. Yet the player makes deductions constantly. A name, phrase, object, or contradiction in one clip can suggest a search term. If that term finds new clips, the case moves forward. When it works, it creates one of the strongest eureka feelings in detective games.

Search systems also help with following leads, a part of detective work many games automate. In Blackwell, interesting place names heard in conversation or seen in text can be typed into an in-game internet search to reveal addresses, which then become locations on the map. A name becomes useful only when the player has noticed that it is useful.

Sherlock Holmes: Consulting Detective uses a similar principle with a newspaper and a directory. The player scans articles for names, businesses, and details that may matter to the case, then looks up possible leads among a large set of entries. Relevant names are hidden in a sea of red herrings, so the player must read carefully and decide what deserves attention.

Once the detective has the facts, the case needs an accusation. The weak version is simply walking up to a person and accusing them. Some Assassin's Creed investigations work like this: if the player chooses the wrong person, they lose some points and try someone else. If they choose correctly, the case ends. That does not ask the player to prove they understand what happened.

A better accusation system requires the player to show their work. Eagle Eye Mysteries, a surprisingly effective DOS mystery game for children, asks the player to gather many statements and then, when ready, pick the key statements plus the suspect's profile. Because the player has many statements and must choose several important ones, they need a real understanding of the crime rather than a lucky guess.

Return of the Obra Dinn takes another route. The player uses a pocket watch on corpses to enter frozen moments of death, then works out who each person was and how they died. The Crew Muster Roll lets the player write down those deductions using names and fate entries. There are far too many combinations to guess through. The player has to know the answer before the game will accept it.

The Vanishing of Ethan Carter uses a timeline puzzle, asking the player to number ghostly images in the correct order to reconstruct a murder. Even multiple-choice questionnaires can work better at the end of a case, as in Consulting Detective, because the player must enter the courtroom confident in their theory of the murderer and motive.

Detective games are a fascinating design challenge because the fantasy is intellectual. Detectives are admired because they know which questions to ask, which lead to chase, how to connect dots, and how to see the facts from a different angle. Games can let the player feel that process more directly than any passive medium can.

Yet many detective games weaken that promise by holding the player's hand and almost giving away the answer. If the interface asks the decisive question, shows the possible answers, or turns every thought into an obvious puzzle piece, the player may solve the interface rather than the mystery.

The strongest systems test the player's intelligence without accidentally prompting them. They let the player gather information, notice contradictions, follow leads, form deductions away from the interface, search for names and ideas they personally spotted, and finally prove the accusation with evidence. A good detective game does not just ask the player to choose the right answer. It gives them enough room to have the thought in the first place.