Walking and Deducing — The Boundary from Gone Home to Return of the Obra Dinn

Introduction

You enter an empty house, open drawers, and read notes left on desks. For the first dozen minutes of Gone Home (2013), that is all that happens. No enemies, no timer, not even a concept of failure. And yet, as you pass through one room after another, an image of what befell a family slowly rises. Is this a 'puzzle'? I treat that question as a boundary you cannot avoid when thinking about the design of thinking games.

At one end sits the walking simulator, where a story reaches you simply by walking and reading the environment. At the other sits the deduction puzzle, where you actively combine gathered fragments and are made to enter an explicit 'answer.' The two get lumped together as 'games where you explore to learn a story,' but their skeletons differ completely. Here I line up Gone Home and Firewatch against Return of the Obra Dinn and The Case of the Golden Idol to dig into where the design fault line runs. This is a second axis, distinct from the one I drew in narrative and storyless puzzles.

The Reading Pole — What a Walking Simulator Designs

February 2012's Dear Esther is often cited as the origin of this lineage. All you can do is walk a deserted Hebridean island and hear fragmentary narration. No branches, no explicit puzzles. Yet the order and pace of your steps are left to you, and at the moment the landscape and the narration overlap, an image of a story about a lost wife coheres. What is designed here is not a 'solution' but an 'allocation of attention.' Terrain and light gently steer where, and in what order, you look.

August 2013's Gone Home condensed that method into a single house. Walking a 1995 Oregon home, reading letters, ticket stubs, and diary pages, you watch a sister's story surface. February 2016's Firewatch carries a two-hander over the single thread of a radio, and April 2017's What Remains of Edith Finch stages each family member's end as a short string of inputs. What they share is that the player has almost no room to 'get it wrong.' The order may shuffle, but the image always coheres in the end.

So at this pole difficulty never stands in front of you. There is no unclimbable vertical wall of the sort I discussed in shaping the learning curve. What the designer tunes is the order and density of the information handed out, and the length of the suspended time in which 'the whole picture isn't visible yet.' The player's hand looks passive, but is really reading the world through a single verb: looking. This pole shares a border with the works in turning looking into play.

Deduction as Action — Assembling an Answer from Fragments

By contrast, October 2018's Return of the Obra Dinn does not advance a step if you only walk and read. You board a ghost ship, observe frozen tableaux of the moment of death, and enter, one by one, the identity and fate of sixty crew members into a logbook. Only when three entries are correct does the confirmation chime sound. There is a clear 'entry of an answer,' and a verdict of right or wrong. What decisively separates it from a walking simulator is that a moment is designed in which the player actively declares to the world, 'this is the answer.'

October 2022's The Case of the Golden Idol laid this structure even barer. You extract proper nouns and verbs from each scene and fit words into gapped sentences to assemble the truth of an incident. What you gather are fragments of information; what you solve is the work of ordering those fragments into the correct causality. Where a walking simulator says 'read it and you will understand,' this says 'gather all you like—without assembling it, you will not.' That assembly step is the very core that makes deduction puzzles an active kind of play.

Activity comes at a cost. That a player can declare an answer means a player can be wrong. Here the judgment I discussed in the ethics of undo—how far to forgive trial—comes into force. Obra Dinn reveals correctness only in units of three, thwarting brute force. Golden Idol quietly accepts a wrong answer and lets you rebuild as often as you like. The deduction puzzle must design how to handle a state that never existed in a walking simulator: failure. Here lies the first fault line dividing the two poles.

Where the Boundary Lies — Is There a Round Trip of 'Solved'

So where is the line drawn? I take 'is there an explicit entry of an answer' as the first criterion. Gone Home has no input field; reach the last room and the story simply completes. Obra Dinn has its logbook, Golden Idol its fill-in blanks. Whether there is a round trip in which the player returns 'this is how I read it' to the world and the world answers 'correct / not yet.' The presence of that round trip is the backbone separating environmental narrative from the deduction puzzle.

Yet this boundary is not as sharp as it sounds. Many works deliberately try to stand right on the line. In the next section I want to step into that blurred territory.

The Blended Zone — The Compromises of Her Story and Outer Wilds

June 2015's Her Story stands exactly on the line. The player digs through a database of interview footage using only search terms. Searching is an active act, yet there is no mechanism to judge correctness. The moment of feeling you have 'understood' the story exists only inside the player. I treated Sam Barlow's design thinking separately as the philosophy of discarding the container; he deliberately omits the answer field and still makes active deduction happen. It sits exactly between the walking simulator's passivity and the deduction puzzle's activity.

2019's Outer Wilds shows a different compromise. The player roams a solar system freely, reading the environment—close, in that sense, to a walking simulator. But the knowledge gathered is auto-recorded in a ship's computer, and in the end the player must go to the right place in the right order or the story will not complete. There is no answer field, but there is an 'execution of the answer' using the body. That opinion on this work splits is likely because those expecting passive exploration and those expecting active puzzle-solving enter through the same door.

May 2024's Lorelei and the Laser Eyes may be the latest form of this compromise. Roaming a mansion makes it a walking simulator; entering the right number into a lock makes it unmistakably a deduction puzzle. Reading the environment and entering an answer become two faces of a single action. The boundary has not vanished; it has been internalized within one work. The designer is no longer asked 'which pole to build,' but 'how to blend both poles within one work.'

Closing

The pole of walking and reading, and the pole of active deduction. What separates them is neither genre name nor look, but a single point: whether there is a round trip in which the player returns an answer to the world. The lineage of environmental narrative refined from Dear Esther through Gone Home and Firewatch, and the lineage of deduction puzzles crystallized from Obra Dinn to Golden Idol, melt quietly together in the middle ground of Her Story, Outer Wilds, and Lorelei. It is not a matter of which is better. Only the experience sold differs: the passive cohering of an image, or the active pleasure of assembly.

If I built something in this territory next, I would settle one thing first: will I make the player 'enter an answer,' or not? If yes, I must at the same time design how to receive their mistakes. If no, I must guarantee, through terrain and light alone, an order in which the image always coheres. Aiming for the middle is alluring, but as the split verdict on Outer Wilds shows, the danger of betraying a passivity-seeker with activity always follows. I leave the reader a question: the last time you felt you had 'solved' a narrative game, was it the moment you returned an answer to the world, or the moment an image cohered on its own?