Her Story is a search box

Her Story, Sam Barlow, 2015, hands you a police database and a search box. The clips of a woman being interviewed are all already there. Nothing is gated behind a cutscene you have not earned, except your own ignorance of which word to type. The obvious object is a mystery plot. The actual object is a search.

That is why it feels like detection. You decide what matters, at your own pace, and a wrong word is not a fail state. It is an empty result, which is information. A mystery that leads you from clue to clue has already done the thinking. A mystery that waits for you to name the next term lets you be wrong in a way you can learn from. The plot is what you assemble. The system is the box.

Searching in Her Story feels like detection because the game hands you the whole case and lets you decide what matters next, at whatever pace you want.

What you are digging through is old police interview footage. Hannah Smith talks about Simon, her murdered husband. You never watch those interviews straight through, though: each one has been chopped into single answers and scattered across a wheezing old database called LOGIC.

Retrieving a clip means typing a word or phrase into the terminal. Any clip in which someone says that word comes back in the results list.

Described flatly, that is a search box. Played, it is hard to put down. Hannah says Glasgow, so Glasgow goes into the box. That turns up a name, so the name goes in next. Then a place, an object, a phrase she keeps returning to, a detail that does not square with an earlier answer.

You start hearing motifs. You take notes, sketch timelines, follow guesses, and work out which trails are dead ends. The haze lifts a little at a time, and the story lands as something personal, because you were the one who put it together.

Her Story sidesteps the hardest part of interrogation design by casting you as the listener rather than the questioner. The starting point was the police interview itself: that room as a small stage where a detective leans on a suspect, steers them wrong, and waits for the crack to show.

Plenty of games have chased that scene, and conversation keeps refusing to be simulated. In Ace Attorney the player is frequently a beat ahead of the interviewer. LA Noire narrows you to a short menu of questions and replies, which can feel like picking from a list rather than investigating.

Genuine conversational freedom is close to impossible. Aisle, a text game, accepts nearly any command you can think of, but only because the whole thing stops after one command. A fine trick for a miniature, no use across a full mystery.

So Her Story flips the seat. The questions were asked years ago and are already on tape. Your work is to comb the archive, chase the answers, and pay attention to what the suspect repeats, sidesteps, and contradicts.

The testimony became a web rather than a sequence because the clues sit in the language of the answers, not behind chapters that unlock. A more conventional build nearly happened instead: hold parts of the testimony back until the player had watched particular clips or entered exactly the right keywords.

Reading real interview transcripts pointed somewhere better. In one murder case, answers that were nominally about hobbies and daily routine kept drifting back to money. Search the transcript for cash-related words and you could leap around it, pulling a hidden motive out of replies that looked entirely harmless.

That is the idea Her Story was built on. Give the player the run of the archive from the first minute, and let the language itself carry the leads.

In the finished game any word you like runs against the whole stack of 271 clips. Nothing is sealed off. What keeps the mystery intact is the writing, the ordering, the ambiguity, and the missing context, not a locked door.

The opening searches are a tutorial in disguise: the early clips show you how to listen, what counts as a lead, and how to build a theory out of fragments. Freedom does not mean being dropped in with nothing, and there is a reasonably clear road out, starting with a word already sitting in the search box: "murder".

A clip tagged with that word sends you to Simon. Simon sends you to another subject, and that to a name, a place, a relationship, another thread. Half a dozen searches in, you have a rough map and several directions worth trying.

What matters is that none of it reads as a corridor. Most clips carry more than one searchable idea. Some of those pay off, some are noise, and some only make sense much later.

Call it guided freedom. You are steering, and you can feel it, while the early material quietly does the teaching.

The archive is written to answer the hunches players actually have, so a good guess pays out and a lazy one is politely turned away.

Type in something broad and unsupported, a stab in the dark about a knife, and you may get back a mild clip about cutting someone's hair with a bread knife. The search worked. The result tells you the theory did not.

Catch a bruise or a tattoo, though, and search around it, and the clips that come back feel like the game winking at you. You noticed something, you tested it, and a fresh seam of testimony opened up.

That is the moment the game catches fire. A tattoo gets described as an apple and a snake. Simon guesses a name from the image. Eve occurs to you, you type it, more clips arrive. It feels earned because it started with something you observed, not something the game flagged.

Occasionally the anticipation runs the other way and takes a theory off the table. Start suspecting a twin and there is a clip waiting to argue against it. The database seems responsive because the script already knows which questions players tend to ask.

Capping results at five is what stops an open archive from spilling its ending on your third search. If everything is reachable from the start, the obvious risk is that somebody types the right word early and pulls a late, revealing clip straight out of the pile.

The rule that handles it is one of the game's neatest. Whatever you type, you see only the first five matching clips, ordered by where they fall in the database.

The most loaded testimony sits late, and the earlier interviews have already spent many of the same words on duller business. Mirror, body, and baby all matter, but tamer early clips tend to fill the visible slots before anything dangerous surfaces.

Nothing in that cheats. The late clips are never locked, and in principle you can reach them whenever you like. The author simply gets to weight the order, so most players meet the story in an order that works.

A revelation only spoils the story if you have the context to decode it, and Her Story withholds that context on purpose. The writing is the second line of defense: stumble onto a significant late clip and it may still slide past you.

Someone describes what happened using nothing sturdier than "she" and "her". Until you know who those pronouns point at, the clip is not the smoking gun it looks like.

Held on its own, every scrap of testimony stays slippery. Two women, tangled relationships, names that recur, deliberate misdirection, and daylight between what is said and what is true. One clip is likelier to raise a question than settle one.

And the game is not purely a whodunnit. It is a whydunnit. Picking a killer out of a lineup interests it much less than working out why any of this happened, and how much of Hannah's account you are willing to believe.

Handing the player the wheel is a craft job: freedom only reads as freedom when the designer has shaped the space it happens in. Her Story lets you take a story apart on your own terms, and shows there is real technique in letting go of someone's hand.

Open-ended is not the same as unattended. The work is still there. Plant strong opening terms. Load clips with more than one lead. Pay off observation, push back on premature theories, keep the result window narrow. And make sure a lone revelation stays ambiguous until enough context has landed around it.

From the inside it feels like wandering an archive under your own steam. Underneath, the game is nudging you toward productive ground and keeping you from wrecking the mystery early. You only confirm that balance by playing the build and watching where curiosity goes.

That is why it holds up. It trusts you without handing over authorship. Search terms become clues, curiosity becomes structure, and a heap of unconnected clips becomes a mystery you believe you solved.

In the Flockbay app, write six short clips and a search that returns them by words you choose. Do not add a highlight on the next word. Play until you are stuck, and notice which word you wish you had tried. If you cannot get stuck, the game is leading you.

A detective mystery template and the AI point-and-click maker fit a single computer. Six clips are enough.